DESCRIPTION
TRANSMISSION DEVICE, RECEPTION DEVICE, AND OFDM TRANSMISSION
METHOD
[Technical Field]
[0001]
The present invention relates to a data transmission
technique using OFDM (Orthogonal Frequency Division
Multiplexing) with pilot signals scattered on a carrier-symbol
plane.
[Background Art]
[0002]
OFDM is a transmission method according to which a large
number of mutually orthogonal carriers aremodulated with digital
data to be transmitted and the resulting modulated waves are
multiplexed and transmitted. With an increase in the number
of carriers used, the symbol duration increases as compared with
a single-carrier transmission method at the same transmission
rate. This makes the OFDM transmission more robust to the effect
of multi-path propagation. In OFDM, in addition, a redundant
duration called a guard interval is inserted between adjacent
symbols to avoid the inter-symbol interference that multi-path
propagation would cause.
[0003]
However, in a multi-path, the phases and amplitudes of the
individual carries vary, so that the receiver needs to compensate
(equalize) the distortion in phase and amplitude. In one scheme

employed widely for the equalization, pilot signals (of which
phase and amplitude are known to the receiver) are transmitted
in some of the cells (i.e., the transmission units identified
by combinations of the symbol number and the carrier number)
contained in an OFDM signal. The receiver estimates the channel
characteristics by using the pilot signals and equalizes the
received signal by using the estimated channel characteristics .
[0004]
For example, according to DVB-T (Digital Video Broadcasting
Terrestrial) system, which is the standard for the broadcast
transmission of digital terrestrial television in Europe, and
ISDB-T (Integrated Services Digital Broadcasting Terrestrial)
system in Japan, pilot signals called SPs (Scattered Pilots)
are scattered on a carrier-symbol plane (hereinafter referred
to as "k-n plane" ) in a pattern shown in FIG. 17 (See Non-Patent
Documents 1 and 2 listed below) . In each figure showing a signal
arrangement pattern of SP signals, the vertical axis represents
a time axis and "n" represents a symbol number, whereas the
horizontal axis represents a frequency axis, and k represents
a carrier number. In addition, an open circle represents a cell
carrying an SP signal, whereas a black dot represents a cell
carrying data. Data referred herein includes data representing
video andaudio information andalso include control information,
such as TPS (Transmission Parameter Signaling) in DVB-T and TMCC
(Transmission Multiplexing Configuration Control) in ISDB-T.
In each figure showing a signal arrangement pattern of SP signals
on the k-nplane, the symbol numbers starts from 0 and the carrier
numbers starts from 0.

[0005]
Here, let Tu represents the useful symbol duration, Tg
represents the guard interval duration, and Ts (= Tu + Tg)
represents the symbol duration. Then, the interval between
adjacent cells in the same carrier in the direction of the time
axis is equal to Ts, and the interval between adjacent cells
in the same symbol in the direction of the frequency axis is
equal to 1/Tu.
[0006]
In FIG. 17, SP signals in each symbol appear at an interval
of 12 carriers and SP signals in each carrier appear at an interval
of 4 symbols . The position of each SP signal is shifted by three
carriers per symbol. That is, when kSP(n) denotes the carrier
number of a cell containing an SP signal in a symbol having the
symbol number n, this carrier number kSp(n) satisfies the
following Equation 1, where mod is a modulus operator and p is
an integer greater than or equal to 0.
[0007]
[Equation 1]

Each SP signal is modulated based on the pseudo-random
binary sequence wk, and the amplitude and phase of the SP signal
is determined depending exclusively on the carrier number k of
the cell containing that SP signal and not on the symbol number
n.
[0008]

With reference to FIG. 18, the following describes the
principles of the channel estimation and equalization performed
by the receiver by using SP signals . FIG. 18 is a block diagram
showing the structure of a typical receiver.
[0009]
In a receiver 100, a signal received with a non-illustrated
receiving antenna is subjected to predetermined processes by
non-illustrated components including a tuner. The processed
signal is supplied to a Fourier transform unit 101 where signal
parts each containing a useful symbol duration Tu are clipped
from the supplied signal and the Fourier transform is applied
to the clipped signal parts to convert the clipped signal parts
into a reception signal Y' (n, k) . The converted reception signal
Y'(n, k) is output to a division unit 106 and also to an SP
extraction unit 102. The SP extraction unit 102 extracts a
reception SP signal Y' (n, kSp(n)) from the reception signal Y' (n,
k) and outputs the extracted reception SP signal Y' ( n, kSp (n) )
to a division unit 104.
[0010]
An SP generation unit 103 generates a nominal SP signal
Y(n, kSP(n)), which is identical to an SP signal generated by
the transmitter, and outputs the SP signal Y(n, kSP(n)) to the
division unit 104. The division unit 104 divides the reception
SP signal Y' (n, kSP(n) ) by the SP signal Y(n, kSP(n) ) and outputs
the division result as the channel response H'(n, kSP(n)) to
an interpolation unit 105. The interpolation unit 105
interpolates the channel response H' (n, kSP(n) ) on the k-n plane
to estimate the channel response H' (n, k) for each cell and outputs

the thus estimated channel response H'(n, k) to the division
unit 106.
[0011]
The division unit 106 divides the reception signal Y' (n,
k) by the channel response H' (n, k) to estimate a transmission
signal X'(n, k) and outputs the thus estimated transmission
signal X'(n, k).
[0012]
Through the above processes, the distortion in amplitude
and phase of the transmission signal caused by multi-path is
compensated using SP signals (See Patent Document 1, for
example).
[0013]
In addition, disclosed is the application of MIMO (Multiple
Input Multiple Output) techniques, which employ multiple
antennas at both the transmitter and receiver to achieve
high-speed and high-capacity data transmission, to a digital
terrestrial television broadcasting using SP signals, such as
DVB-T (See Non-Patent Document 3, for example).
[0014]
First, the following describes the overview of a MIMO
transmission system in which the transmitter and the receiver
both have two antennas, with reference to FIG. 19. FIG. 19 is
a diagram showing such a MIMO transmission system.
[0015]
A transmitter 200 transmits a first transmission signal
and a second transmission signal from a first transmitting
antenna 2 01 and a second transmitting antenna 2 02 , respectively.

The first transmission signal is obtained by applying the inverse
Fourier transform to a first transmission signal Xc1 (n, k) , and
the second transmission signal is obtained by applying the
inverse Fourier transform to a second transmission signal Xc2 (n,
k) . Note that the first and second transmission signals are
simultaneously transmitted respectively on the cells each having
the symbol number n and the carrier number k.
[0016]
A receiver 300 receives a first reception signal with a
receiving antenna 301. The first reception signal contains the
first transmission signal arrived via a channel Pc11 and the
second transmission arrived via a channel Pcl2. The receiver
300 applies the Fourier transform to the first reception signal
to obtain a first reception signal Yc'1(n, k). In addition,
the receiver 3 00 receives a second reception signal with a
receiving antenna 302. The second reception signal contains
the second transmission signal arrived via a channel Pc22 . The
receiver 300 applies the Fourier transform to the second
reception signal to obtain a second reception signal Yc'2(n,
k) . The receiver 300 then conducts a predetermined process on
the first reception signal Yc'1(n, k) and the second reception
signal Yc'2(n, k) and outputs the first transmission signal
Xc'1(n, k) and the second transmission signal Xc'2(n, k).
[0017]
Here, let He11 (n, k) , Hcl2 (n, k) , Hc21(n, k) , and Hc22 (n,
k) respectively denote the channel responses of channels Pc11,
Pc12, Pc21, and Pc22 at the cell having the symbol number n and
the carrier number k . Let Nc1(n, k) and Nc2 (n, k) denote the

noise power contained in the first reception signal Yc'1(n, k)
and in the second reception signal Yc'2(n, k), respectively.
Then, the first reception signal Yc'1(n, k) and the second
reception signal Yc'2(n, k) are expressed by Equation 2 shown
below. The notation [] in Equation 2 represents a matrix.
[0018]
[Equation 2]

That is, once the channel responses of the channels Pc11,
Pc12, Pc21, and Pc22 are estimated, the receiver 300 is able
to separate and equalize the first transmission signal Xc'l(n,
k) and the second transmission signal Xc'2 (n, k) by using Equation
3 shown below, where Hc' 11 (n, k) , Hc ' 12 (n, k) , Hc' 21 (n, k) , and
Hc'22 (n, k) are the channel responses estimated by the receiver
300. In Equation 3, the notation []represents a matrix, and
the notation []-1 represents the inverse matrix of [].
[0019]
[Equation 3]

Non-Patent Document 3 describes a technique for enabling
separation and estimation of channel responses of two channels
from two transmitting antennas to one receiving antenna, by
transmitting SP signals arranged in the pattern shown in FIG.
17 from the first transmitting antenna and SP signals arranged

in the pattern shown in FIG. 20 from the second transmitting
antenna. In FIG. 20, a plus ( + ) sign indicates that the polarity
of an SP signal transmitted from the second transmitting antenna
is not inverted with respect to the polarity of a corresponding
SP signal transmitted from the first transmitting antenna. On
the other hand, a minus (-) sign indicates that the polarity
of an SP signal transmitted from the second transmitting antenna
is inverted with respect to the polarity of a corresponding SP
signal transmitted from the first transmitting antenna.
[0020]
That is, of the SP signals transmitted from the second
transmitting antenna, the polarity of each SP signal having an
even symbol number is not inverted and of each SP signal having
an odd symbol number is inverted, with respect to the polarity
of a corresponding SP signal transmitted from the first
transmitting antenna.
[0021]
The receiver observes, for each symbol where the symbol
number n is an even number, components representing the sum of
the channel responses of the two channels, one of which is from
the first transmitting antenna to the receiving antenna and the
other is from the second transmitting antenna to the receiving
antenna (hereinafter, the former is referred to as "first channel
response" and the latter as "second channel response") . On the
other hand, for each symbol where the symbol number n is an odd
number, components representing the difference between the first
and second channel responses are observed. Therefore, the
receiver can separate and estimate the first channel response

by adding the sum components and the difference components, and
the second channel response by subtracting the difference
components from the sum components.
[Non-Patent Document 1]
"Digital Video Broadcasting (DVB); Framing structure,
Channel coding and modulation for digital terrestrial
television", ETSI EN 300 744 by European Telecommunications
Standards Institutes
[Non-Patent Document 2]
"TRANSMISSION SYSTEM FOR DIGITAL TERRESTRIAL TELEVISION
BROADCASTING", ARIB STD-B31 by Association of Radio Industries
and Businesses
[Non-Patent Document 3]
"ADUALPOLARIZATIONMIMOBROADCASTTVSYSTEM" , BBCResearch
White Paper WHP 144 by J. D. Mitchell, P. N. Moss andM. J. Thorp
[Patent Document 1]
JP patent No. 2772286
[Summary of Invention]
[Technical Problem]
[0022]
The following now considers the range in which a channel
response is duly estimated on condition that SP signals are
arranged in the pattern shown in FIG. 17, which is used in the
DVB-T system as well as in the ISDB-T system.
[0023]
FIG. 21 is a schematic view of responses on the delay
time-Doppler frequency plane (hereinafter referred to as the

"τ-fD plane") of SP signals arranged on the k-n plane in the
pattern shown in FIG. 17. In other words, FIG. 21 show
two-dimensional Fourier transform pairs of SP signals arranged
on the k-n plane in the pattern shown in FIG. 17 . In each figure
showing SP signal responses and showing the estimatable ranges
of channel responses, the horizontal axis represents a delay
time axis (hereinafter referred to as the "Τ axis") and
corresponds to the delay time (T) of the impulse response of
a channel. The vertical axis represents a Doppler frequency
axis (hereinafter referred to as the "fD axis") and corresponds
to the Doppler frequency (fD) of the Doppler spectrum of a channel.
In addition, a black dot represents a response of an SP signal
on the τ-fD plane.
[0024]
As shown in FIG. 21, the minimum interval between SP signal
responses on the τ-fD plane in the τ axis direction is equal
to Tu/12 . It is because SP signals on the k-n plane are arranged
to appear one for every 12 carriers in the same symbol. In other
words, the sampling interval in the k axis direction is equal
to 12/Tu. Further, the minimum interval between SP signal
responses on the τ-fD plane in the fD axis direction is equal
to 1/(4Ts) . It is because SP signals on the k-n plane are arranged
to appear one for every 4 symbols in the same carrier. In other
words, the sampling interval in the n axis direction is equal
to 4Ts. Still further, the minimum interval between SP signal
responses on the τ-fD plane at the same Doppler frequency in
the i axis direction is equal to Tu/3 . It is because the minimum
interval between SP signals on the k-n plane in the kaxis direction

is equal to 3 carriers. Still further, the minimum interval
between SP signal responses at the same delay time on the τ-fD
plane in the fD axis direction is equal to 1/Ts. It is because
the minimum interval between SP signals on the k-n plane in the
n axis direction is equal to one symbol.
[0025]
In the case where an impulse response of a channel has a
delay spread, the response spreads in the T axis direction as
compared with a corresponding SP signal response. In the case
where a Doppler spectrum of a channel has a frequency spread,
the spectrum spreads in the fD axis direction as compared with
a corresponding SP signal response.
[0026]
FIG. 22 shows a region of the τ-fD plane in which the channel
response H' (n, kSP(n) ) of an SP signal can be interpolated without
causing aliasing distortion, on condition that the channel
response H' (n, kSp(n)) is first interpolated in the n axis
direction and then in the k axis direction of the k-n plane.
In FIG. 22, a black dot represents an SP signal response on the
τ-fD plane, and a rectangle represents a channel response of
the channel from the transmitting antenna to the receiving
antenna.
[0027]
From FIG. 22, it is known that a rectangular region having
a width of Tu/3 in the T axis direction and a width of 1/(4Ts)
in the fD axis direction is the region in which the channel response
is interoperated without causing aliasing distortion
(hereinafter, referred to as "interpolatable region").

According to the DVB-T and ISDB-T systems, the length of the
longest guard interval duration is Tu/4 . With the guard interval
duration equal to Tu/4, the spread of the impulse response of
the channel equal to Tu/4 or less would not adversely affect
the reception quality. It is because the inter-symbol
interference is ensured to fall within the guard interval
duration. The width of the interpolatable region in the τ axis
direction is set to Tu/3 in order to allow a margin for practical
filters and yet to ensure a correct estimation of a channel
response without incurring the risk of inter-symbol
interference.
[0028]
As described above, in terms of the design details of a
transmission system, the guard interval duration and the SP
signal arrangement are closely related. That is, in order not
to impair the tolerance to multi-path delay provided by insertion
of guard interval durations, the minimum interval between SP
signals on the k-n plane in the k axis direction needs to be
shorter than a predetermined interval. In terms of the
transmission efficiency, however, it is desirable to keep to
a minimum the density of SP signals, which do not carry any useful
information. That is, there is a trade-off between the guard
interval duration and the SP signal arrangement.
[0029]
FIG. 23 shows a region of the τ-fD plane in which the channel
responseH' (n, kSP(n) ) of an SP signal can be interpolated without
causing aliasing distortion, on condition that the channel
responseH' (n, kSp(n) ) is interpolated only in the k axis direction

and not in the n axis direction of the k-n plane. In FIG. 23,
a black dot represents an SP signal response on the τ-fD plane,
and a rectangle represents a channel response of the channel
from the transmitting antenna to the receiving antenna.
[0030]
From FIG. 23, it is known that a rectangular region having
a width of Tu/12 in the τ axis direction and a width of 1/Ts
in the fD axis direction is a region in which the channel response
is interoperated without causing aliasing distortion
(hereinafter, referred to as "interpolatable region").
[0031]
The following now considers the range in which channel
response is duly estimated with the SP signal arrangement
disclosed in Non-Patent Document 3, which is used for a MIMO
transmission system.
[0032]
The process of inverting and not inverting the polarity
of SP signals transmitted from the first transmitting antenna
is equivalent to an arithmetic operation of multiplying
individual SP signals transmitted from the first transmitting
antenna, by the complex plane wave expressed by the left side
of Equation 4 shown below. The complex plane wave has an
equi-phase line parallel to the kaxis direction on the k-n plane,
and the cycle in the n axis direction is equal to 2n.
[0033]
[Equation 4]


Note that in Equation 4, the right side is obtained by-
rewriting the left side using the relation n = (1/Ts)t.
[0034]
Accordingly, the response of each SP signal transmitted
from the second transmitting antenna is said to be shifted the
response of a corresponding SP signal transmitted from the first
transmitting antenna, by l/(2Ts) in the fD axis direction on
the τ-fD plane.
[0035]
Inviewof the above, the responses of SP signals transmitted
from the first transmitting antenna and the responses of SP
signals from the second transmitting antenna are expressed on
the same τ-fD plane as shown in FIG. 24. Note that a black dot
represents a response of an SP signal transmitted from the first
transmitting antenna, whereas a cross represents a response of
an SP signal transmitted from the second transmitting antenna.
[0036]
Note that the process of inverting and not inverting the
polarity of SP signals transmitted from the first transmitting
antenna shown in FIG. 2 0 is to invert the polarity of an SP signal
transmitted from the first transmitting antenna at every third
carrier in the frequency direction. In other words, the process
may be construed to be equivalent to an arithmetic operation
of multiplying individual SP signals transmitted from the first
transmitting antenna, by the complex plane wave expressed by
the left side of Equation 5 shown below. The complex plane wave
has an equi-phase line parallel to the n axis on the k-n plane

and the cycle in the k axis direction is equal to 6k.
[0037]
[Equation 5]

Note that in Equation 5, the right side is obtained by-
rewriting the left side using the relation k = Tuf . In addition,
the phase term in Equation 5 is attached with a negative (-)
sign. It is because the delay in the positive direction along
the T axis corresponds to the phase rotation exp(-j2nfi) in the
negative direction in proportion to the frequency f.
[0038]
Based on the above understanding, it is said that the
response of each SP signal transmitted from the second
transmitting antenna is shifted the response of a corresponding
SP signal transmitted from the first transmitting antenna, by
Tu/6 in the τ axis direction on the τ-fDplane. It is thus apparent
from that each response shown in FIG. 2 0 is equivalent to that
obtained by shifting the response of a corresponding SP signal
shown in FIG. 24 by 1/(2Ts) in the fD axis direction.
[0039]
The receiver divides each received SP signal (i.e., a mixed
SP signal which is a mixture of an SP signal transmitted from
the first transmitting antenna and an SP signal transmitted from
the second transmitting antenna) by the nominal SP signal. As
a result of the division, the receiver obtains a channel response
which is a mixture of a channel response of the channel from

the first transmitting antenna to the receiving antenna (the
first channel response) and a channel response of the channel
from the second transmitting antenna to the receiving antenna
(the second channel response).
[0040]
The first channel response has the spreading from the black
dots shown in FIG. 24, in accordance with the impulse response
and Doppler spectrum. Similarly, the second channel response
has the spreading from the crosses shown in FIG. 24, in accordance
with the impulse response and Doppler spectrum.
[0041]
FIG. 25 shows a region of the τ-fD plane in which the first
and second channel responses are interpolated without causing
aliasing distort ion and separated from each other without causing
crosstalk therebetween, on condition that the channel response
of each SP signal is interpolated first in the n axis direction
and then in the k axis direction on the k-n plane. In FIG. 25,
a black dot represents a response of an SP signal transmitted
from the first transmitting antenna, whereas a cross represents
a response of an SP signal transmitted from the second
transmitting antenna. In addition, a rectangular with a solid
line represents the first channel response, whereas a rectangular
with a broken line represents the second channel response.
[0042]
From FIG. 25, it is known that a rectangular region having
a width of Tu/6 in the τ axis direction and a width of 1/(4Ts)
in the fD axis direction is what is hereinafter referred to as
"interpolatable & separable region". In the interpolatable &

separable region, the first and second channel responses are
interoperated without causing aliasing distortion and separated
without causing crosstalk therebetween.
[0043]
FIG. 26 shows a region of the τ-fD plane in which the first
and second channel responses are interpolated without causing
aliasing distort ion and separated from each other without causing
crosstalk therebetween, on condition that the channel response
of each SP signal is interpolated in the k axis direction only
and not in the n axis direction on the k-n plane. In FIG. 26,
a black dot represents a response of an SP signal transmitted
from the first transmitting antenna, whereas a cross represents
a response of an SP signal transmitted from the second
transmitting antenna. In addition, a rectangular with a solid
line represents the first channel response, whereas a rectangular
with a broken line represents the second channel response.
[0044]
From FIG. 26, it is known that a rectangular region having
a width of Tu/12 in the τ axis direction and a width of 1/ (2Ts)
in the fD axis direction is what is hereinafter, referred to
as "interpolatable & separable region" . In the interpolatable
& separable region, the first and second channel responses are
interoperated without causing aliasing distortion and separated
without causing crosstalk therebetween.
[0045]
From a comparison of the interpolatable region shown in
FIG. 22 with the interpolatable & separable region shown in FIG.
25, it is shown that the width Tu/6 of the interpolatable &

separable region in the τ axis direction is a half of the width
Tu/3 of the interpolatable region in the τ axis direction. As
mentioned above, it is preferable to set the τ-axis direction
width in which correct estimation of the first and second channel
responses is ensured in a manner not to impair the tolerance
to multi-path delay provided by insertion of guard interval
durations. However, the SP signal transmission method
described in Non-Patent Document 3 is associated with the
following problem, even without considering any margin to be
allowed for practical filters used for interpolation and
separation. That is, in the case where the guard interval
duration is longer than Tu/6, specifically where the guard
interval duration is equal to Tu/4 for example, the tolerance
to multi-path delay achieved by the insertion of guard intervals
is impaired and thus the first and second channel responses may
not be correctly estimated.
[0046]
In addition, from a comparison of the interpolatable region
shown in FIG. 23 with the interpolatable & separable region shown
in FIG. 26, it is shown that the width 1/ (2Ts) of the interpolatable
& separable region in the fD axis direction is a half of the
width 1/Ts of the interpolatable region in the fD axis direction.
As clarified above, the SP signal transmission method according
to Non-Patent Document 3 has a problem in the ability of following
the time variation of a channel.
[0047]
In view of the problems noted above, the present invention
aims to provide a transmitter, a receiver, and an OFDM

transmission method each of which achieves the following
advantages, in the case where a plurality of pilot signals are
transmitted from a plurality of transmitting antennas. The
transmitter, receiver, and OFDM transmission method according
to the present invention ensure correct estimation of a channel
response involving a delay spread to the comparable to the case
where pilot signals are transmitted from a single transmitting
antenna or ensure the ability to follow the time variation of
a channel to the extent comparable to the case where pilot signals
are transmitted from a single transmitting antenna.
[Solution to Problem]
[0048]
In order to achieve the above aim, a transmitter according
to one aspect of the present invention has first to Mth
transmitting antennas (where M is an integer equal to or greater
than 2) and is for transmitting an OFDM signal obtained by
modulating a plurality of carriers per symbol duration. The
OFDM signal contains pilot signals scattered on a carrier-symbol
plane. On the carrier-symbol plane, k denotes a carrier number,
n denotes a symbol number, ∆k denotes an interval between pilot
signals in a same symbol, ∆n denotes an interval between pilot
signals in a same carrier, and p denotes an integer greater than
or equal to 0. ns and ks each denotes a nonzero integer, and
m denotes an integer satisfying 1 ≤ m ≤ M. When 2 ≤ m ≤ M, neither
(m-1)ns nor (m-1)ks is equal to an integral multiple of M. The
carrier number kP(n) of a carrier that transmits a pilot signal
in a symbol with the symbol number n satisfies Equation 6.

[0049]
[Equation 6]

The transmitter includes a generating unit operable to
generate a plurality of pilot signals as mth pilot signals for
an mth antenna (where 1 ≤ m ≤ M) , such that a phase difference
between a phase of each mth pilot signal and a phase of a reference
pilot signal is equal to a value given by Equation 7.
[0050]
[Equation 7]

The transmitter further includes a transmitter operable
to transmit, from the mth transmitting antenna, an OFDM signal
containing the mth pilot signals generated by the generating
unit.
[Advantageous Effects of Invention]
[0051]
According to the above-described aspects of the present
invention, it is ensured that the responses of pilot signals
appearing at the same Doppler frequency on the τ-fD plane are
all transmitted from only one of the first to Mth transmitting
antennas, and also that the responses of pilot signals appearing
at the same delay time on the τ-fD plane are all transmitted

from only one of the first to Mth transmitting antennas . By virtue
of this, the transmitter transmits pilot signals from the first
to Mth transmitting antennas to ensure that the receiver is able
to estimate the channel responses as long as the delay spread
of the pilot signals is the same level that can be accurately
estimated when pilot signals are transmitted from a single
transmitting antenna, or able to follow the channel's time
variability with the accuracy that would be achieved when pilot
signals are transmitted from a single transmitting antenna.
[Brief Description of Drawings]
[0052]
FIG. 1 is a diagram showing the congiruation of a MIMO
transmission system according to a first embodiment of the
present invention;
FIG. 2 is a schematic view showing the arrangement pattern
and polarity reversal of SP signals transmitted from a second
transmitting antenna 12 shown in FIG. 1;
FIG. 3 is a schematic view showing, ona τ-fDplane, responses
of SP signals transmitted from a first transmitting antenna 11
and the second transmitting antenna 12 shown in FIG. 1;
FIG. 4 is a schematic view showing a region in which the
channel responses are separated and estimated, on condition that
SP signals arranged the patterns shown in FIGs. 17 and 2 are
used;
FIG. 5 is another schematic view showing a region in which
the channel responses are separated and estimated, on condition
that SP signals arranged the patterns shown in FIGs. 17 and 2

are used;
FIG. 6 is a diagram showing the structure of a transmitter
10 shown in FIG. 1;
FIG. 7 is a diagram showing the structure of a receiver
3 0 shown in FIG. 1;
FIG. 8 is a diagram showing the structures of channel
separation & estimation units 35 and 38 shown in FIG. 7;
FIG. 9 is a schematic view showing the arrangement pattern
and polarity reversal of SP signals transmitted from the second
transmitting antenna 12 according to a second embodiment of the
present invention;
FIG. 10 is a schematic view showing, on a τ-fD plane,
responses of SP signals transmitted from the first transmitting
antenna 11 and the second transmitting antenna 12 according to
the second embodiment;
FIG. 11 is a schematic view showing a region in which the
channel responses are separated and estimated, on condition that
SP signals arranged in the patterns shown in FIGs. 17 and 9 are
used;
FIG. 12 is another schematic view showing a region in which
the channel responses are separated and estimated, on condition
that SP signals arranged in the patterns shown in FIGs. 17 and
9 are used;
FIG. 13 is a diagram showing the structure of a transmitter
10a according to the second embodiment;
FIG. 14 is a diagram showing the structure of a receiver
3 0a according to the second embodiment;
FIG. 15 is a diagram showing the structures of channel

separation & estimation units 35a and 38a shown in FIG. 14;
FIG. 16 is a view for illustrating a generalization of the
complex plane waves used in the first and second embodiments;
FIG. 17 is a view showing a pattern of SP signal arrangement
used in a DVB-T or ISDB-T system;
FIG. 18 is a diagram showing the structure of a conventional
receiver, for illustrating the principles of cannel estimation
and reception signal equalization performed by the receiver;
FIG. 19 is a diagram showing the structure of a conventional
MIMO transmission system, for illustrating the principles of
a MIMO transmission method;
FIG. 2 0 is a schematic view showing the arrangement pattern
and polarity reversal of SP signals transmitted from a second
transmitting antenna 12 according to a conventional technique;
FIG. 21 is a schematic view showing, on a τ-fD plane, SP
signals arranged in the pattern shown in FIG. 17;
FIG. 22 is a schematic view showing a region in which the
channel responses are separated and estimated, on condition that
SP signals arranged the pattern shown in FIG. 17 are used;
FIG. 23 is another schematic view showing a region in which
the channel responses are separated and estimated, on condition
that SP signals arranged the pattern shown in FIG. 17 are used;
FIG. 24 is a schematic view showing, on a τ-fD plane,
responses of SP signals transmitted from a first transmitting
antenna and a second transmitting antenna according to a
conventional technique;
FIG. 25 is a schematic view showing a region in which the
channel responses are separated and estimated, on condition that

SP signals arranged in the patterns shown in FIGs. 17 and 20
are used; and
FIG. 2 6 is another schematic view showing a region in which
the channel responses are separated and estimated, on condition
that SP signals arranged in the patterns shown in FIGs. 17 and
20 are used.
[Reference Sings List]
[0053]
10 Transmitter
11 First Transmitting Antenna
12 Second Transmitting Antenna
13 SP Generation Unit
14 Complex Plane Wave Generation Unit
15 Multiplication Unit

30 Receiver
31 First Receiving Antenna
32 Second Receiving Antenna
35, 38 Channel Separation & Estimation Unit
51, 61 SP Extraction Unit
52 SP Generation Unit
53, 63 Division Unit
54, 57, 64, 67 Interpolation Unit
55 Complex Plane Wave Generation Unit
56, 66 Multiplication Unit
[Description of Embodiments]
[0054]

One aspect of the present invention provides a first
transmitter having first to Mth transmitting antennas (where
M is an integer equal to or greater than 2) and for transmitting
an OFDM signal obtained by modulating a plurality of carriers
per symbol duration,. The OFDM signal contains pilot signals
scattered on a carrier-symbol plane. On the carrier-symbol
plane, k denotes a carrier number, n denotes a symbol number,
∆k denotes an interval between pilot signals in a same symbol,
∆n denotes an interval between pilot signals in a same carrier,
and p denotes an integer greater than or equal to 0. ns and
ks each denotes a nonzero integer, and m denotes an integer
satisfying 1 ≤ m ≤ M. When 2 ≤ m ≤ M, neither (m-1)ns nor (m-1)ks
is equal to an integral multiple of M. The carrier number kP(n)
of a carrier that transmits a pilot signal in a symbol with the
symbol number n satisfies Equation 8.
[0055]
[Equation 8]

The first transmitter includes a generating unit operable
to generate a plurality of pilot signals as mth pilot signals
for an mth antenna (where 1 ≤ m ≤ M) , such that a phase difference
between a phase of each mth pilot signal and a phase of a reference
pilot signal is equal to a value given by Equation 9.
[0056]
[Equation 9]


The first transmitter further includes a transmitter
operable to transmit, from the mth transmitting antenna, an OFDM
signal containing the mth pilot signals generated by the
generating unit.
[0057]
According to the above-described aspects of the present
invention, it is ensured that the responses of pilot signals
appearing at the same Doppler frequency on the τ-fD plane are
all transmitted from only one of the first to Mth transmitting
antennas, and also that the responses of pilot signals appearing
at the same delay time on the τ-fD plane are all transmitted
from only one of the first to Mth transmitting antennas . By virtue
of this, the first transmitter transmits pilot signals from the
first to Mth transmitting antennas in a manner that the receiver
is able to estimate the channel responses involving a delay spread
to the extent that would be expected when pilot signals are
transmitted from a single transmitting antenna or to follow the
channel's time variability with the accuracy that would be
achieved when pilot signals are transmitted from a single
transmitting antenna.
[0058]
Another aspect of the present invention provides a second
transmitter consistent with the first transmitter, wherein the
generating unit includes: a reference signal generating unit
operable to generate a plurality of first pilot signals, each

first pilot signal being the reference pilot signal; and a
multiplication unit operable to generate the mth pilot signals
where 2 ≤m≤M, each mth pi lot signal being generated by multiplying
the reference pilot signal by a complex plane wave expressed
by Equation 10 on the carrier-symbol plane.
[0059]
[Equation 10]

[0060]
Yet another aspect of the present invention provides a third
transmitter consistent with the first transmitter, wherein the
generating unit includes: a reference signal generating unit
operable to generate a plurality of first pilot signals, each
first pilot signal being the reference signal; and a phase
rotation unit operable to generate the mth pilot signals where
2 ≤ m ≤ M, each mth pilot signal being generated by rotating
the phase of the reference pilot signal by the value given by
Equation 9.
[0061]
Yet another aspect of the present invention provides a
fourth transmitter consistent with the first transmitter,
wherein M is equal to 2. The generating unit includes: a
reference signal generatingunit operable to generate aplurality
of first pilot signals, each first pilot signal being the
reference signal; and a polarity inversion unit operable to
generate aplurality of secondpilot signals such that polarities

of the second pilot signals are each alternately inverted and
not inverted in a carrier direction with respect to a polarity
of a corresponding reference signal in a same symbol.
[0062]
According to the above-described aspects of the present
invention, a means for readily generating the first to mth sets
of pilot signals is provided.
[0063]
Yet another aspect of the present invention provides a fifth
transmitter consistent with the first transmitter, wherein M
is equal to 2, An is equal to 4, Ak is equal to 12, ns is equal
to 1, and ks is equal to 1.
[0064]
Yet another aspect of the present invention provides a sixth
transmitter consistent with the first transmitter, wherein M
is equal to 2, ∆n is equal to 4, ∆k is equal to 12, ns is equal
to 1, and ks is equal to -3.
[0065]
The above-described aspects of the present invention are
directly applicable, for example, to the DVB-T system or the
ISDB-T system.
[0066]
Yet another aspect of the present invention provides a first
receiver for receiving an OFDM signal transmitted from a
transmitter having a plurality of first to Mth transmitting
antennas (where M is an integer greater than or equal to 2).
The OFDM signal is obtained by modulating a plurality of carriers
per symbol duration. The OFDM signal contains a plurality of

pilot signals scattered on a carrier-symbol plane. On the
carrier-symbol plane, k denotes a carrier number, n denotes a
symbol number, ∆k denotes an interval between pilot signals in
a same symbol, ∆n denotes an interval between pilot signals in
a same carrier, and p denotes an integer greater than or equal
to 0. ns and ks each denotes a nonzero integer, and m denotes
an integer satisfying 1 ≤ m ≤ M. When 2 ≤ m ≤ M, neither (m-1)ns
nor (m-1)ks is equal to an integral multiple of M. The carrier
number kP(n) of a carrier that transmits a pilot signal in a
symbol with the symbol number n satisfies Equation 11.
[0067]
[Equation 11]

A plurality of pilot signals transmitted from an mth one
of the transmitting antennas (where m is an integer satisfying
1 ≤ m ≤ M) are mth pilot signals, such that a phase difference
between a phase of each mth pilot signal and a phase of a reference
pilot signal is equal to a value given by Equation 12.
[0068]
[Equation 12]

The first receiver includes : a receiving antenna with which
the OFDM signal from the transmitter is received; a response
estimation unit operable to estimate a channel response of each

of first to Mth channels respectively from the first to Mth antennas
to the receiving antenna, the estimation being carried out based
on Equation 12 and pilot signals contained in the OFDM signal
received with the receiving antenna; and a signal estimation
unit operable to estimate first to Mth transmission signals based
on the received OFDM signal and the estimated channel responses
of the first to Mth transmission channels, the first to Mth
transmission signals corresponding to first to Mth OFDM signals
transmitted respectively from the first to Mth transmitting
antennas.
[0069]
According to the above-described aspects of the present
invention, it is ensured that the responses of pilot signals
appearing at the same Doppler frequency on the τ-fD plane are
all transmitted from only one of the first to Mth transmitting
antennas, and also that the responses of pilot signals appearing
at the same delay time on the τ-fD plane are all transmitted
from only one of the first toMth transmitting antennas . By virtue
of this, when pilot signals are transmitted from the first to
Mth transmitting antennas, the receiver is able to estimate the
channel responses involving a delay spread to the extent that
wouldbe expected whenpilot signals are transmitted from a single
transmitting antenna or to follow the channel' s time variability
with the accuracy that would be achieved when pilot signals are
transmitted from a single transmitting antenna.
[0070]
Yet another aspect of the present invention provides a
second receiver consistent with the first receiver, wherein the

response estimation unit is operable to extract pilot signals
from the OFDM signal received with the receiving antenna, divide
each extracted pilot signal by the reference pilot signal, and
estimate the channel response of the first channel based on a
result of each division, and further operable to estimate the
channel response of mth channel where 2 ≤ m ≤ M, based on the
result of each division and Equation 12.
[0071]
According to the above-described aspect of the present
invention, the channel response of the mth channel is readily
estimated.
[0072]
Yet another aspect of the present invention provides a firs t
OFDM transmission method for transmitting an OFDM signal from
a transmitter having first to Mth transmitting antennas (where
M is an integer greater than or equal to 2) . The OFDM signal
is obtained by modulating a plurality of carries per symbol
duration. The OFDM signal contains pilot signals scattered on
a carrier-symbol plane. On the carrier-symbol plane, k denotes
a carrier number, n denotes a symbol number, ∆k denotes an interval
between pilot signals in a same symbol, ∆n denotes an interval
between pilot signals in a same carrier, and p denotes an integer
greater than or equal to 0. ns and ks each denotes a nonzero
integer, and m denotes an integer satisfying 1 ≤ m ≤ M. When
2 ≤ m ≤ M, neither (m-1)ns nor (m-1)ks is equal to an integral
multiple of M. The carrier number kP(n) of a carrier that
transmits a pilot signal in a symbol with the symbol number n
satisfies Equation 13 .

[0073]
[Equation 13]
Ak
kp(n) = — x(n modAn) + Akxp
An
The first OFDM transmission method includes the step of
generating a plurality of pilot signals as mth pilot signals
for an mth antenna (where 1 ≤ m ≤ M) , such that a phase difference
between a phase of each mth pilot signal and a phase of a reference
pilot signal is equal to a value given by Equation 14.
[0074]
[Equation 14]

The first OFDM transmission method further includes the
step of transmitting, from the mth transmitting antenna, an OFDM
signal containing the mth pilot signals generated in the
generating step.
[0075]
According to the above-described aspects of the present
invention, it is ensured that the responses of pilot signals
appearing at the same Doppler frequency on the τ-fD plane are
all transmitted from only one of the first to Mth transmitting
antennas, and also that the responses of pilot signals appearing
at the same delay time on the τ-fD plane are all transmitted
from only one of the first to Mth transmitting antennas . By virtue

of this, the transmitter transmits pilot signals from the first
to Mth transmitting antennas in a manner that the receiver is
able to estimate the channel responses involving a delay spread
to the extent that would be expected when pilot signals are
transmitted from a single transmitting antenna or to follow the
channel's time variability with the accuracy that would be
achieved when pilot signals are transmitted from a single
transmitting antenna.
[0076]
The following describes embodiments of the present
invention, with reference to the accompanying drawings.
[0077]
<>
First, a first embodiment of the present invention is
described with reference to the drawings. In the following
description, a useful symbol duration is denoted by Tu, and a
guard interval duration is denoted by Tg, and a symbol duration
is denoted by Ts (= Tu + Tg) , similarity to the description of
a conventional examples given above.
[0078]

A MIMO transmission system according to the present
embodiment is described with reference to FIG. 1. FIG. 1 is
a diagram showing the congiruation of the MIMO transmission
system according to the present embodiment.
[0079]
The MIMO transmission system 1 shown in FIG. 1 includes:

a transmitter 10 having two transmitting antennas 11 and 12;
and a receiver 30 having two receiving antennas 31 and 32.
[0080]
The transmitter 10 transmits a first transmission signal
and a second transmission signal from a first transmitting
antenna 11 and a second transmitting antenna 12, respectively.
The first transmission signal is obtained by applying the inverse
Fourier transform to a first transmission signal X1(n, k) ,
whereas the second transmission signal is obtained by applying
the inverse Fourier transform to a second transmission signal
X2 (n, k) . The first and second transmission signals are
simultaneously transmitted respectively on the cells each having
a symbol number n and a carrier number k. Note that each the
first and second transmission signals transmitted from the first
and second transmitting antenna 11 and 12 is a signal generated
by modulating multiple mutually orthogonal carriers with data
to be transmitted as well as with other data, followed by
multiplexing of the resultant modulated waves.
[0081]
The receiver 30 receives with the receiving antenna 31 a
first reception signal containing the first transmission signal
arrived via a channel P11 and the second transmission signal
arrived via a channel P12 and applies the Fourier transform to
the first reception signal to obtain a first reception signal
Y'l(n, k). In addition, the receiver 30 receives with the
receiving antenna 32 a second reception signal containing the
first transmission signal arrived via a channel P21 and the second
transmission signal arrived via a channel P22 and applies the

Fourier transform to the second reception signal to obtain a
second transmission signal Y'2(n, k). The receiver 30 then
conducts a predetermined process on the first reception signal
Y1' (n, k) and the second reception signal Y2 ' (n, k) and outputs
a first transmission signal X1' (n, k) and a second transmission
signal X2'(n, k).
[0082]

Before the detailed description of the transmitter 10 and
the receiver 30 shown in FIG. 1, a description is given of the
description of SP signals transmitted from the first transmitting
antenna 11 and SP signals transmitted from the second
transmitting antenna 12.
[0083]
The SP signals transmitted from the first transmitting
antenna 11 are arranged in the pattern shown in FIG. 17 mentioned
above. The value of the complex number of each SP signal
allocated to a cell is equal to the complex number of a typical
SP signal allocated to a corresponding cell in the DVB-T and
ISDB-T systems.
[0084]
In contrast, the SP signals transmitted from the second
transmitting antenna 12 are arranged in the pattern shown in
FIG. 2. Note that a plus ( + ) sign in FIG. 2 indicates that the
polarity of each SP signal transmitted from the second
transmitting antenna 12 in a cell represented by a plus ( + ) sign
is not inverted with respect to the polarity of an SP signal
transmitted from the first transmitting antenna 11 in a

corresponding cell having the same symbol number and the same
carrier number. On the other hand, a minus (-) sign in FIG.
2 indicates that the polarity of each SP signal transmitted from
the second transmitting antenna 12 in a cell represented by a
minus (-) sign is inverted with respect to the polarity of an
SP signal transmitted from the first transmitting antenna 11
in a corresponding cell with the same symbol number and the same
carrier number.
[0085]
As shown in FIGs . 17 and 2 , the SP signals transmitted from
the first transmitting antenna 11 are arranged in the same pattern
as the SP signals transmitted from the second transmitting
antenna 12. In addition, both the patterns of SP signal
arrangement satisfy that the carrier number kSP(n) of a cell
transmitting an SP signal in the symbol having the symbol number
n satisfies Equation 1 described above.
[0086]
The SP signals transmitted from the second transmitting
antenna 12 in one symbol are alternately inverted andnot inverted
in polarity, with respect to the SP signals transmitted from
the first transmitting antenna 11 in the same symbol. In the
direction that the symbol number is incremented by 1 and the
carrier number is incremented by 3, the polarity of SP signals
transmitted from the second transmitting antenna 12 are all
inverted or non-inverted with respect to the polarity of
corresponding SP signals transmitted from the first transmitting
antenna 11.
[0087]

The process of inverting or not inverting the polarity of
SP signals transmitted from the first transmitting antenna 11
is equivalent to an arithmetic operation of multiplying
individual SP signals transmitted from the first transmitting
antenna 11 by the complex plane wave expressed by the left side
of Equation 15 shown below. The complex plane wave has a cycle
equal to 8n in the n axis direction and a cycle equal to 24k
in the k axis direction on the k-n plane.
[0088]
[Equation 15]

Note that in Equation 15, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0089]
Accordingly, the response of each SP signal transmitted
from the second transmitting antenna 12 is said to be shifted
the response of a corresponding SP signal to be transmitted from
the first transmitting antenna 11, byTu/24 in the T axis direction
and l/(8Ts) in the fD axis direction on the τ-fD plane.
[0090]
Inviewof the above, the responses of SP signals transmitted
from the first transmitting antenna 11 and the responses of SP
signals transmitted from the second transmitting antenna 12 are
expressed on the same τ-fD plane as shown in FIG. 3. Note that
a black dot in FIG. 3 represents a response of an SP signal

transmitted from the first transmitting antenna 11, whereas a
cross represents a response of an SP signal transmitted from
the second transmitting antenna 12.
[0091]
The receiver 3 0 divides each received SP signal by the
nominal SP signal to obtain a mixed channel response which is
a mixture of a channel response of the channel from the first
transmitting antenna 11 to one of the receiving antennas (either
the first receiving antenna 31 or the second receiving antenna
32) and a channel response of the channel from the second
transmitting antenna 12 to the one of the receiving antennas
(the former channel response is referred as the "channel response
related to the first transmitting antenna 11" and the latter
is referred to as the "channel response related to the second
transmitting antenna 12").
[0092]
However, the channel response related to the first
transmitting antenna 11 has the spreading from the black dots
shown in FIG. 3, in accordance with the impulse response and
Doppler spectrum. Similarly, the channel response related to
the second transmitting antenna 12 has the spreading from the
crosses shown in FIG. 3, in accordance with the impulse response
and Doppler spectrum.
[0093]
FIG. 4 shows a region of the τ-fD plane in which the channel
response related to the first transmitting antenna 11 and the
channel response related to the second transmitting antenna 12
are interpolated without causing aliasing distortion and

separated from each other without causing crosstalk therebetween,
on condition that the channel response of SP signals are
interpolated first in the n axis direction and then in the k
axis direction on the k-nplane. In FIG. 4, a black dot represents
a response of an SP signal transmitted from the first transmitting
antenna 11, whereas a cross represents of an SP signal transmitted
from the second transmitting antenna 12. In addition, a
rectangular with a solid line represents the channel response
related to the first transmitting antenna 11, whereas a
rectangular with a broken line represents the channel response
related to the second transmitting antenna 12.
[0094]
From FIG. 4, it is known that a rectangular region having
a width of Tu/3 in the τ axis direction and a width of l/(8Ts)
in the fD axis direction is what is hereinafter refereed to as
an "interpolatable & separable region" . In the interpolatable
& separable region, the channel response related to the first
transmitting antenna 11 and the channel response related to the
second transmitting antenna 12 are interoperated without causing
aliasing distortion andseparated from eachotherwithout causing
crosstalk.
[0095]
From a comparison of FIG. 4 with FIG. 22, the following
is noted on condition that SP signals transmitted from the first
transmitting antenna 11 are arranged in the pattern shown in
FIG. 17 and that SP signals transmitted from the second
transmitting antenna 12 are arranged in the pattern shown in
FIG. 2. Here, the width of the interpolatable & separable region

in the τ axis direction is Tu/3, which is equal to the width
of the interpolatable region in the τ axis direction shown in
FIG. 17. As mentioned above, the interpolatable region shown
in FIG. 17 is a region in which the channel response of SP signals
are transmitted from a single transmitting antenna is
interoperated without causing aliasing distortion. As
described above, the width in the τ axis direction in which correct
estimation of both the channel responses related to the first
and second transmitting antennas 11 and 12 is ensured is set
so as not to impair the tolerance to multτ-path delay provided
by insertion of guard interval durations.
[0096]
FIG. 5 shows a region of the τ-fD plane in which the channel
response related to the first transmitting antenna 11 and the
channel response related to the second transmitting antenna 12
are interpolated without causing aliasing distortion and
separated from each other without causing crosstalk therebetween,
on condition that the channel response of SP signals are
interpolated only in the k axis direction and not in the n axis
direction on the k-n plane. In FIG. 5, a black dot represents
a response of an SP signal transmitted from the first transmitting
antenna 11, whereas a cross represents a response of an SP signal
transmitted from the second transmitting antenna 12. In
addition, a rectangular with a solid line represents the channel
response related to the first transmitting antenna 11, whereas
a rectangular with a broken line represents the channel response
related to the second transmitting antenna 12.
[0097]

From FIG. 5, it is known that a rectangular region having
a width of Tu/24 in the τ axis direction and a width of 1/Ts
in the fD axis direction is what is hereinafter referred to as
an "interpolatable & separable region" . In the interpolatable
& separable region, the channel response related to the first
transmitting antenna 11 and the channel response related to the
second transmitting antenna 12 are interoperated without causing
aliasing distortion and separated from each other wi thout caus ing
crosstalk therebetween.
[0098]
From a comparison of FIG. 5 with FIG. 23, the following
is noted on condition that SP signals transmitted from the first
transmitting antenna 11 are arranged in the pattern shown in
FIG. 17 and that SP signals transmitted from the second
transmitting antenna 12 are arranged in the pattern shown in
FIG. 2. Here, the width of the interpolatable & separable region
in the fD axis direction is 1/Ts, which is equal to the width
of the interpolatable region in the fD axis direction shown in
FIG. 17. As mentioned above, the interpolatable region shown
in FIG. 17 is a region in which the channel response of SP signals
transmitted from a single transmitting antenna is interoperated
without causing aliasing distortion. That is, the fD
axis-direction width of a region in which correct estimation
is ensured for both the channel response related to the first
transmitting antenna 11 and the channel response related to the
second transmitting antenna 12 is not impaired at all. In other
words, the ability of following the channel's time variability-
is not impaired.

[0099]
Further, FIG. 24 and FIG. 3 are compared.
[0100]
In FIG. 24, the responses appearing at the same Doppler
frequency are a mixture of the responses of SP signals transmitted
from the first transmitting antenna and the responses of SP
signals transmitted from the second transmitting antenna.
Therefore, the impulse responses of the respective channels share
the same region in the x axis direction. Also, the responses
appearing at the same delay time are a mixture of the responses
of SP signals transmitted from the first transmitting antenna
and the responses of SP signals transmitted from the second
transmitting antenna. Therefore, the Doppler spectrums of the
respective channels share the same region in the fD axis
direction.
[0101]
In FIG. 3, in contrast, the responses appearing at the same
Doppler frequency are exclusively of the responses of SP signals
transmitted from either of the first transmitting antenna 11
and the second transmitting antenna 12 . Therefore, the impulse
responses of the respective channels are allowed to occupy
mutually different regions in the τ axis direction. Also, the
responses appearing at the same delay time are exclusively of
the responses of SP signals transmitted from either of the first
transmitting antenna 11 and the second transmitting antenna 12 .
Therefore, the Doppler spectrums of the respective channels are
allowed to occupy mutually different regions in the fD axis
direction.

[0102]
The difference noted above is derived from the difference
in the shift direction and shift amount of SP signals on the
τ-fD plane, in other words from the difference in the direction
and frequency of the equτ-phase lines of SP signals on the k-n
plane.
[0103]
As described above, according to the present embodiment
that uses the two SP signal arrangements shown in FIGs. 17 and
2, the interpolatable & separable region is extended widthwise
in the τ or fD axis direction, as compared with the interpolatable
& separable region in the conventional case where the two SP
signal arrangements shown in FIGs. 17 and 20 are used. As
described above, the interpolatable & separable region refers
to a region in which the channel response related to the first
transmitting antenna 11 and the channel response related to the
second transmitting antenna 12 are interpolated without causing
aliasing distortion andseparated from eachotherwithout causing
crosstalk therebetween.
[0104]

With reference to FIG. 6, the following describes the
structure of the transmitter 10 shown in FIG. 1. FIG. 6 is a
diagram showing the structure of the transmitter 10 shown in
FIG. 1.
[0105]
As described above, the transmitter 10 has the first
transmitting antenna 11 and the second transmitting antenna 12 .

Additionally, the transmitter 10 has an SP generation unit 13,
a complex plane wave generation unit 14, a multiplication unit
15, a mapping unit 16, a cell allocation unit 17, an inverse
Fourier transform unit 18, a guard interval insertion unit 19,
a mapping unit 20, a cell allocation unit 21, an inverse Fourier
transform unit 22, and a guard interval insertion unit 23.
[0106]
[SP Generation Unit 13]
The SP generation unit 13 generates SP signals modulated
based on the pseudo-random binary sequence wk. The SP generation
unit 13 outputs the generated SP signals to the cell allocation
unit 17 and also to the multiplication unit 15. The SP signals
output to the cell allocation unit 17 are used as SP signals
to be transmitted from the first transmitting antenna 11. The
amplitude and phase of each SP signal is determined only by the
carrier number k of the cell to which the SP signal is allocated
and thus independently of the symbol number n.
[0107]
[Complex Plane Wave Generation Unit 14]
The complex plane wave generation unit 14 generates a
complex plane wave expressed by Equation 16 below and outputs
the generated complex plane wave to the multiplication unit 15 .
[0108]
[Equation 16]

In Equation 16, n represents the symbol number and k

represents the carrier number.
[0109]
[Multiplication Unit 15]
The multiplication unit 15 multiplies an SP signal received
from the SP generation unit 13 by the complex plane wave received
from the complex plane wave generation unit 14 and outputs the
result of the multiplication to the cell allocation unit 21 where
the result of the multiplication is used as an SP signal to be
transmitted from the second transmitting antenna 12 . Note that
the process of themultiplicationperformed by themultiplication
unit 15 is equivalent to the process of shifting the SP signal
received from the SP generation unit 13, by Tu/24 in the τ axis
direction and by 1/(8Ts) in the fD axis direction on the τ-fD
plane. It goes without saying that an SP signal and a complex
plane wave used in a multiplication of that SP signal are of
a pair having the same symbol number and the same carrier number.
[0110]
[Mapping Unit 16]
The mapping unit 16 receives data on which a predetermined
process has been conducted. The mapping unit 16 maps data
composed of a predetermined number of bits at a predetermined
bit unit onto a predetermined constellation and outputs data
resulting from the mapping to the cell allocation unit 17.
[0111]
Note that the mapping units 16 and 2 0 may use any of various
types of constellations, such as constellations for PSK (Phase
Shift Keying) , QAM (Quadrature Amplitude Modulation) , and APSK
(Amplitude Phase Shift Keying).

[0112]
[Cell Allocation Unit 17]
The cell allocation unit 17 allocates SP signals input from
the SP generation unit 13 (i.e., SP signals transmitted from
the first transmitting antenna 11) to cells specif ied by Equation
1 mentioned above, and allocates data input from the mapping
unit 16 to the other cells. As a result of the allocation in
this manner, the cell allocation unit 17 constitutes frames and
output a frame signal relating to the frames to the inverse Fourier
transform unit 18.
[0113]
[Inverse Fourier Transform Unit 18]
The inverse Fourier transform unit 18 applies, symbol by
symbol, the inverse Fourier transform to the modulated symbols
(i.e. , data output from the mapping unit 16 and SP signals output
from the SP generation unit 13) allocated to the cells
corresponding to the carriers included in the frame signal input
from the cell allocation unit 17, thereby to carry out the OFDM
modulation. As a result of the OFDM modulation, the inverse
Fourier transform unit 18 generates a modulated signal having
a useful symbol duration Tu into which multiple mutually
orthogonal carriers are modulated and multiplexed. The
modulated signal having the useful symbol duration Tu is then
output to the guard interval insertion unit 19.
[0114]
[Guard Interval Insertion Unit 19]
The guard interval insertion unit 19 generates a signal
having a guard interval duration Tg (hereinafter, the signal

is referred to as "guard interval signal" ) , based on the modulated
signal of the useful symbol duration Tu input from the inverse
Fourier transform unit 18. The guard interval insertion unit
19 then inserts the guard interval signal to the modulated signal
having the useful symbol duration Tu and outputs the resulting
modulated signal (hereinafter, referred to as "OFDM signal") .
The OFDM signal output from the guard interval insertion unit
19 is subjected to predetermined processes, including
digital-to-analog conversion, frequency conversion to the
transmission band, and amplification, and then transmitted as
the first transmission signal from the first transmit ting antenna
11.
[0115]
Note that each guard interval signal inserted to the
modulated signal by the guard interval insertion units 19 and
23 corresponds to a guard interval part of a signal obtained
by cyclically and continually repeating a modulated signal of
the useful symbol duration Tu. In one specific example given
for the purpose of description, the guard interval signal is
a signal having the duration Tg and identical to a latter part
of the modulation signal of the useful symbol duration Tu.
Alternatively, the guard interval signal may be a signal
modulatedbased on a predeterminedpseudo-randombinary sequence
or a zero-amplitude signal.
[0116]
[Mapping Unit 20]
The mapping unit 20 receives data on which a predetermined
process has been conducted. The mapping unit 20 maps data

composed of a predetermined number of bits at a predetermined
bit unit onto a predetermined constellation and outputs data
resulting from the mapping to the cell allocation unit 21.
[0117]
[Cell Allocation Unit 21]
The cell allocation unit 21 allocates the multiplication
result input from the multiplication unit 15 (i.e., an SP signal
transmitted from the second transmitting antenna 12) to cells
specified by Equation 1 mentioned above, and data input from
the mapping unit 2 0 to the other cells. As a result of the
allocation in this manner, the cell allocation unit 21
constitutes frames and output a frame signal relating to the
frames to the inverse Fourier transform unit 22.
[0118]
[Inverse Fourier Transform Unit 22]
The inverse Fourier transform unit 22 applies, symbol by
symbol, the inverse Fourier transform to the modulated symbols
(i.e. , data output from the mapping unit 2 0 and the multiplication
result output from the multiplication unit 15) allocated to the
cells corresponding to the carriers included in the frame signal
input from the cell allocation unit 21, thereby to carry out
the OFDM modulation. As a result of the OFDM modulation, the
inverse Fourier transform unit 22 generates a modulated signal
having a useful symbol duration Tu into which multiple mutually
orthogonal carriers are modulated and multiplexed. The
modulated signal having the useful symbol duration Tu is then
output to the guard interval insertion unit 23.
[0119]

[Guard Interval Insertion Unit 23]
The guard interval insertion unit 23 generates a signal
having a guard interval duration Tg (hereinafter, the signal
is referred to as "guard interval signal" ) , based on the modulated
signal having the useful symbol duration Tu input from the inverse
Fourier transform unit 22. The guard interval insertion unit
23 then inserts the guard interval signal to the modulated signal
having the useful symbol duration Tu and outputs the resulting
modulated signal (hereinafter, referred to as "OFDM signal").
The OFDM signal output from the guard interval insertion unit
23 is subjected to predetermined processes, including
digital-to-analog conversion, frequency conversion to the
transmission band, and amplification, and then transmitted as
the second transmission signal from the second transmitting
antenna 12 as the second transmission signal.
[0120]

Withreference to FIG. 6, the followingdescribes operations
of the transmitter having the structure described above.
[0121]
The SP generation unit 13 generates SP signals and outputs
the generated SP signals to the cell allocation unit 17 and also
to the multiplication unit 15. The SP signals output to the
cell allocation unit 17 are used as SP signals to be transmitted
from the first transmitting antenna 11. In addition, the complex
plane wave generation unit 14 generates the complex plane wave
expressed by Equation 16 and outputs the generated complex plane
wave to the multiplication unit 15. The multiplication unit

15 multiplies each SP signal received from the SP generation
unit 13 by the complex plane wave received from the complex plane
wave generation unit 14 and outputs the result of the
multiplication to the cell allocation unit 21 where the received
multiplication result is used as an SP signal to be transmitted
from the second transmitting antenna 12.
[0122]
The mapping unit 16 conducts the mapping process on the
input data and outputs the mapped data to the cell allocation
unit 17. The cell allocation unit 17 organizes frames by-
allocating SP signals input from the SP generation unit 13 (i.e.,
an SP signal to be transmitted from the first transmitting antenna
11) as well as data input from the mapping unit 16 to cells
constituting the frames. The cell allocation unit 17 then
outputs a frame signal relating to the frames to the inverse
Fourier transform unit 18. The inverse Fourier transform unit
18 applies the inverse Fourier transform symbol by symbol to
the frame signal received from the cell allocation unit 17,
thereby generating a modulated signal of each useful symbol
duration Tu. The guard interval insertion unit 19 inserts a
signal of the guard interval duration Tg (guard interval
signal)to the modulated signal of each useful symbol duration
Tu to generate a modulated signal (OFDM signal) in which the
symbol duration Tu and the guard interval signal repeatedly
appear. The guard interval insertion unit 19 then outputs the
OFDM signal, which is then subjected to a predetermined process
and transmitted from the first transmitting antenna 11 as the
first transmission signal.

[0123]
In parallel with the above processing, the following
processing is also performed.
[0124]
The mapping unit 20 conducts a mapping process on the input
data and outputs the mapped data to the cell allocation unit
21. The cell allocation unit 21 organizes frames by allocating
SP signals input from the SP generation unit 15 (i.e., an SP
signal transmitted from the second transmitting antenna 12) as
well as data input from the mapping unit 20 to cells constituting
the frame. The cell allocation unit 20 outputs a frame signal
relating to the frames to the inverse Fourier transform unit
22. The inverse Fourier transform unit 22 applies the inverse
Fourier transform symbol by symbol to the frame signal received
from the cell allocation unit 21, thereby generating a modulated
signal of each useful symbol duration Tu. The guard interval
insertion unit 23 inserts a signal of a guard interval duration
Tg (guard interval signal) to the modulated signal of each useful
symbol duration Tu to generate a modulated signal (OFDM signal)
in which the symbol duration Tu and the guard interval signal
repeatedly appear. The guard interval insertion unit 23 then
outputs the OFDM signal, which is then subjected to a predetermine
process and transmitted from the second transmitting antenna
12 as the second transmission signal. Note that the second
transmission signal is output in synchronism with the first
transmission signal to ensure that cells having the same symbol
number and the same carrier number of the two transmission signals
are simultaneously transmitted.

[0125]

With reference to FIG. 7, the following describes the
structure of the receiver 30 shown in FIG. 1. FIG. 7 is a diagram
showing the structure of the receiver 30 shown in FIG. 1.
[0126]
As described above, the receiver 30 has the first receiving
antenna 31 and the second receiving antenna 32. Additionally,
the receiver 30 has a guard interval removable unit 33 , a Fourier
transform unit 34, a channel separation & estimation unit 35,
a guard interval removable unit 36, a Fourier transform unit
37, a channel separation & estimation unit 38, and a signal
separation & equalization unit 39.
[0127]
[Guard Interval Removable Unit 33]
The first receiving antenna 31 receives a first reception
signal from the transmitter 10. The first reception signal is
a mixed signal of the first transmission signal (OFDM signal)
transmitted from the first transmitting antenna 11 and the second
transmission signal (OFDM signal) transmitted from the second
transmitting antenna 12 . The first reception signal received
with the first receiving antenna 31 is subjected to a
predetermined process and then input to the guard interval
removable unit 33.
[0128]
The guard interval removable unit 33 removes the guard
interval durations from the received first reception signal and
outputs the first reception signal remaining after the removable

of the guard interval durations to the Fourier transform unit
34.
[0129]
[Fourier Transform Unit 34]
The Fourier transform unit 34 receives the first reception
signal from the guard interval removable unit 33, . The Fourier
transform unit 34 applies the Fourier transform to each signal
part carrying a useful symbol duration Tu thereby to carry out
the OFDM demodulation (frequency separation) and outputs the
first reception signal Y'1(n, k) obtained as a result of the
OFDM demodulation to the signal separation & equalization unit
39 and also to channel separation & estimation unit 35.
[0130]
[Channel Separation & Estimation Unit 35]
The channel separation & estimation unit 35 separates and
estimates the channel response H'11(n, k) of the channel P11
and the channel response H'12(n, k) of the channel P12, with
the use of the first reception signal Y'1(n, k) received from
the Fourier transform unit 34. The channel separation &
estimation unit 35 then outputs the channel responses H'11(n,
k) and H'12(n, k) to the signal separation & equalization unit
39.
[0131]
[Guard Interval Removable Unit 36]
The second receiving antenna 32 receives a second reception
signal from transmitter 10. The second reception signal is a
mixed signal of the first and second transmission signals
transmitted from the first and second transmitting antennas 11

and 12 of the transmitter 10, respectively. The second reception
signal received with the second receiving antenna 32 is subjected
to a predetermined process and then input to the guard interval
removable unit 36.
[0132]
The guard interval removable unit 36 removes guard interval
durations from the received second reception signal and outputs
the second reception signal remaining after the removable of
guard interval durations to the Fourier transform unit 37.
[0133]
[Fourier Transform Unit 37]
The Fourier transform unit 37 receives the second reception
signal from the guard interval removable unit 36. The Fourier
transform unit 37 applies the Fourier transform to each signal
part carrying a useful symbol duration Tu thereby to carry out
the OFDM demodulation. The Fourier transform unit 37 then
outputs the second reception signal Y' 2 (n, k) obtained as a result
of the OFDM demodulation to the signal separation & equalization
unit 39 and also to channel separation & estimation unit 38.
[0134]
[Channel Separation & Estimation Unit 38]
The channel separation & estimation unit 38 separates and
estimates a channel response H'21(n, k) of the channel P21 and
the channel response H'22 (n, k) of the channel P22 with the sue
of the second reception signal Y' 2 (n, k) received from the Fourier
transform unit 37. The channel separation & estimation unit
38 then transmits the estimated channel responses H'21(n, k)
and H'22 (n, k) to the signal separation & equalization unit 39.

[0135]
[Signal Separation & Equalization unit 39]
The signal separation & equalization unit 39 receives the
first reception signal Y'1 (n, k) from the Fourier transform unit
34 and also receives the second reception signal Y'2 (n, k) from
the Fourier transform unit 37. In addition, the signal
separation & equalization unit 39 receives the channel responses
H'll(n, k) andH'12(n, k) from the channel separation & estimation
unit 35 and also receives the channel response H'21(n, k) and
H'22(n, k) from the channel separation & estimation unit 38.
[0136]
The signal separation & equalization unit 39 computes
Equation 17 shown below, with the first reception signal Y' 1 (n,
k), the second reception signal Y'2(n, k), and the channel
responses H'11 (n, k) , H'12(n, k) , H'21(n, k) , and H'22(n, k) ,
thereby separating and equalizing the first transmission signal
X'l(n, k) and the second transmission signal X'2(n, k). The
signal separation & equalization unit 39 then outputs the first
transmission signal X'1 (n, k) and the second transmission signal
X'2(n, k) . In Equation 17, notation [] represent a matrix and
the notation []-1 represents the inverse matrix of [].
[0137]
[Equation 17]

The following describes the details of the channel

separation & estimation units 35 and 38 shown in FIG. 7, with
reference to FIG. 8. FIG. 8 is a diagram showing the structures
of the channel separation & estimation units 35 and 3 8 shown
in FIG. 7.
[0138]
[Channel Separation & Estimation Unit 35]
The channel separation & estimation unit 35 includes an
SP extraction unit 51, an SP generation unit 52, a division unit
53, an interpolation unit 54, a complex plane wave generation
unit 55, a multiplication unit 56, and an interpolation unit
57.
[0139]
(SP Extraction Unit 51)
With the use of Equation 1 described above, the SP extraction
unit 51 extracts a first reception SP signal Y'l(n, kSP(n) ) from
the first reception signal Y'1(n, k) that is received from the
Fourier transform unit 34 . The SP extraction unit 51 then outputs
the extracted first reception SP signal Y'l( n, ksp(n) ) to the
division unit 53.
[0140]
(SP Generation Unit 52)
The SP generation unit 52 generates a nominal SP signal
Y(n, kSP(n) ) , which is identical to the SP signal generated by
the SP generation unit 13 of the transmitter 10. The SP
generation unit 52 then outputs the generated nominal SP signal
Y(n, kSP(n)) to the division unit 53 and also to a later-described
division unit 63, which is included in the channel separation
& estimation unit 38.

[0141]
(Division Unit 53)
The division unit 53 multiples the first reception SP signal
Y'l(n, kSP(n) ) by the SP signal Y(n, kSP(n)) and outputs the
division result (hereinafter referred to as the "first mixed
channel response") to the interpolation unit 54 and the
multiplication unit 56.
[0142]
(Interpolation Unit 54)
The interpolation unit 54 is provided with a low-pass filter.
With the use of the low-pass filter, the interpolation unit 54
removes the channel response related to the second transmitting
antenna 12 from the first mixed channel response that is received
from the division unit 53 , and interpolates the channel response
related to the first transmitting antenna 11 on the k-n plane.
In this way, the channel response H'11(n, k) at every cell is
estimated. Then, the interpolation unit 54 outputs the
estimated channel response H'11(n, k) to the signal separation
& equalization unit 39. Note that the low-pass filter included
in the interpolation unit 54 may be, for example, a low-pass
filter having a transfer function from -Tu/6 to Tu/6 in the τ
axis direction and from -1/(16Ts) to 1/(16Ts) in the fD axis
in the direction (see FIG. 4) or a low-pass filter with having
a transfer function from -Tu/48 to Tu/48 in the τ axis direction
and from -1/(2Ts) to 1/(2Ts) in the fD axis in the direction
(see FIG. 5) . In practice, it is desirable to employ a low-path
filter having a bandwidth determined in a manner of allowing
for margin of error expected in practical filters.

[0143]
(Complex Plane Wave Generation Unit 55)
The complex plane wave generation unit 55 generates a
complex plane wave expressed by Equation 18 shown below and
outputs the generated complex plane wave to the multiplication
unit 56 and a later-described multiplication unit 66 of the
channel separation & estimation unit 38.
[0144]
[Equation 18]

Note in Equation 18, n represents the symbol number and
k represents the carrier number.
[0145]
(Multiplication Unit 56)
The multiplication unit 56 multiplies the division result
received from the division unit 53 (the first mixed channel
response) by the complex plane wave received from the complex
plane wave generation unit 55 and outputs the result to the
multiplication (hereinafter referred to as the "first
shif ted-and-mixed channel response" ) to the interpolation unit
57. Note that the multiplication performed by the
multiplication unit 56 is equivalent to the process of shifting
the first mixed channel response by -Tu/24 in the T axis direction
and by -l/(8Ts) in the fD axis direction on the τ-fD plane. It
goes without saying that an SP signal and a complex plane wave
used in a multiplication of that SP signal are of a pair having

the same symbol number and the same carrier number.
[0146]
(Interpolation Unit 57)
The interpolation unit 57 is provided with a low-pass filter.
With the use of the low-pass filter, the interpolation unit 57
removes the channel response related to the first transmitting
antenna 11 from the first shifted-and-mixed channel response
that is received from the multiplication unit 56, and
interpolates the channel response related to the second
transmitting antenna 12 on the k-n plane. In this way, the
channel response H'12(n, k) at every cell is estimated. Then,
the interpolation unit 57 outputs the estimated channel response
H'12(n, k) to the signal separation & equalization unit 39 . Note
that the low-pass filter included in the interpolation unit 57
may be, for example, a low-pass filter having the same a transfer
function as that of the low-pass filter included in the
interpolation unit 54.
[0147]
[Channel Separation & Estimation Unit 38]
The channel separation & estimation unit 38 includes an
SP extraction unit 61, the division unit 63, an interpolation
unit 64, the multiplication unit 66, and an interpolation unit
67. Note that the channel separation & estimation unit 38 may
additionally include components equivalent to the SP generation
unit 52 and the complex plane wave generation unit 55.
[0148]
(SP Extraction Unit 61)
With the use of Equation 1 described above, the SP extraction

unit 61 extracts a second reception SP signal Y'2( n, KSp (n) )
from the second reception signal Y'2 (n, k) that is received from
the Fourier transform unit 37. The SP extraction unit 61 then
outputs the second reception SP signal Y'2( n, ksp(n) ) to the
division unit 63.
[0149]
(Division Unit 63)
The division unit 63 multiples the second reception SP
signal Y'2(n, kSP(n) ) by the SP signal Y(n, kSP(n) ) and outputs
the division result (hereinafter referred to as the "secondmixed
channel response") to the interpolation unit 64 and the
multiplication unit 66.
[0150]
(Interpolation Unit 64)
The interpolation unit 64 is provided with a low-pass filter.
With the use of the low-pass filter, the interpolation unit 64
removes the channel response related to the second transmitting
antenna 12 from the secondmixed channel response that is received
from the division unit 63 , and interpolates the channel response
related to the first transmitting antenna 11 on the k-n plane.
In this way, the channel response H'12(n, k) at every cell is
estimated. Then, the interpolation unit 64 outputs the
estimated channel response H'12(n, k) to the signal separation
& equalization unit 39. Note that the low-pass filter included
in the interpolation unit 64 may be, for example, a low-pass
filter having the same a transfer function as that of the low-pass
filter included in the interpolation unit 54.
[0151]

(Multiplication Unit 66)
The multiplication unit 66 multiples the division result
(the second mixed channel response) received from the division
unit 63, by the complex plane wave received from the complex
plane wave generation unit 55 and outputs the result of the
multiplication (hereinafter referred to as the "second
shifted-and-mixed channel response") to the interpolation unit
67. Note that the multiplication performed by the
multiplication unit 66 is equivalent to the process of shifting
the secondmixed channel response by-Tu/24 in the T axis direction
and-l/(8Ts) in the fD axis direction on the τ-fD plane. It goes
without saying that a secondmixed channel response and a complex
plane wave used in a multiplication of that second mixed channel
response are of a pair having the same symbol number and the
same carrier number.
[0152]
(Interpolation Unit 67)
The interpolation unit 67 is provided with a low-pass filter.
With the use of the low-pass filter, the interpolation unit 67
removes the channel response related to the first transmitting
antenna 11 from the second shifted-and-mixed channel response
that is received from the multiplication unit 66 and interpolates
the channel response related to the second transmitting antenna
12 on the k-n plane. In this way, the channel response H'22 (n,
k) at every cell is estimated. Then, the interpolation unit
67 outputs the estimated channel response H'22 (n, k) to the signal
separation & equalization unit 39 . Note that the low-pass filter
included in the interpolation unit 67 may be, for example, a

low-pass filter having the same a transfer function as that of
the low-pass filter included in the interpolation unit 54.
[0153]

The following describes operations of the receiver having
the above-described structure, with reference to FIGs. 7 and
8.
[0154]
The firs t reception signal received with the first receiving
antenna 31 is subjected to the predetermine process and then
input to the guard interval removable unit 33 where guard interval
durations are removed. After the guard interval removable by
the guard interval removable unit 33, the first reception signal
is input to the Fourier transform unit 34 where the Fourier
transform is applied symbol by symbol to the first reception
signal from which guard interval durations have been removed.
As a result, each signal part carrying a useful symbol duration
Tu is converted into the first reception signal Y'l(n, k)and
the first reception signal Y'l(n, k) is output to the signal
separation & equalization unit 39 and also to the channel
separation & estimation unit 35.
[0155]
In the channel separation & estimation unit 35, a nominal
SP signal Y(n, kSP(n)) is generated by the SP generation unit
52, and a complex plane wave expressed by Equation 18 described
above is generated by the complex plane wave generation unit
55.
[0156]

The SP extraction unit 51 extracts the first reception SP
signal Y'1( n, kgp(n) ) from the first reception signal Y'1(n,
k) received from the Fourier transform unit 34 and outputs the
extracted first reception SP signal Y' 1 ( n, kgp(n) ) to the division
unit 53 . The division unit 53 divides the first reception SP
signal Y'1(n, kSP(n)) received from the SP extraction unit 51,
by the SP signal Y(n, kSP(n)) generated by the SP generation
unit 52. The division result (the first mixed channel response)
is output to the interpolation unit 54 and the multiplication
unit 56.
[0157]
The interpolation unit 54 conducts the above-described
process on the first mixed channel response received from the
division unit 53 , thereby estimating the channel response H' 11 (n,
k) to all the cells and outputs the estimated channel response
H'11(n, k) to the signal separation & equalization unit 39.
[0158]
The multiplication unit 56 multiplies the division result
(the first mixed channel response) received from the division
unit 53, by the complex plane wave received from the complex
plane wave generation unit 55 and outputs the result of the
multiplication (the first shifted-and-mixed channel response)
to the interpolation unit 57 . The interpolation unit 57 conducts
the above-described process on the first shifted-and-mixed
channel response received from the multiplication unit 56,
thereby estimating the channel response H'12 (n, k) at every cell.
The estimated channel response H'12 (n, k) is output to the signal
separation & equalization unit 39.

[0159]
In parallel with the above processing, the following
processing is performed.
[0160]
The second reception signal received with the second
receiving antenna 32 is subjected to the predetermined process
and then input to the guard interval removable unit 3 6 where
guard interval durations are removed. After the guard interval
removable by the guard interval removable unit 36, the second
reception signal is input to the Fourier transform unit 37 where
the Fourier transform is applied symbol by symbol to the second
reception signal from which guard interval durations have been
removed. As a result, each signal part carrying a useful symbol
duration Tu is converted into the second reception signal Y'2 (n,
k) and the second reception signal Y'2(n, k) is output to the
signal separation & equalization unit 39 and also to the channel
separation & estimation unit 38.
[0161]
In the channel separation & estimation unit 38, the SP
extraction unit 61 extracts the second reception SP signal Y'2 (n,
kSP(n)) from the second reception signal Y'2 (n, k) received from
the Fourier transform unit 37 and outputs the extracted second
reception SP signal Y'2 (n, kSP(n)) to the division unit 63 . The
division unit 63 divides the second reception SP signal Y'2 (n,
kSP(n) ) received from the SP extraction unit 61, by the SP signal
Y(n, kSp(n) ) generatedby the SP generation unit 52 . The division
result (the second mixed channel response) is output to the
interpolation unit 64 and also to the multiplication unit 66.

[0162]
The interpolation unit 64 conducts the above-described
process on the second mixed channel response received from the
division unit 63 , thereby estimating the channel response H' 21 (n,
k) at every cell and outputs the estimated channel responseH' 21 (n,
k) to the signal separation & equalization unit 39.
[0163]
The multiplication unit 66 multiplies the division result
received from the division unit 63 (the second mixed channel
response), by the complex plane wave generated by the complex
plane wave generation unit 55 and outputs the result of the
multiplication (the second shif ted-and-mixed channel response)
to the interpolation unit 67 . The interpolation unit 67 conducts
the above-described process on the second shifted-and-mixed
channel response received from the multiplication unit 66,
thereby estimating the channel responseH' 22 (n, k) at every cell.
The estimated channel response H'22 (n, k) is output to the signal
separation & equalization unit 39.
[0164]
The signal separation & equalization unit 39 computes
Equation 17 described above, with the first reception signal
Y'1(n, k) and the second reception signal Y'2(n, k) that are
received as a result of the above two processes performed in
parallel, and also with the channel responses H' 11 (n, k) , H' 12 (n,
k) , H'21(n, k) , andH'22(n, k) , thereby separating and equalizing
the first transmission signal X'1(n, k) and the second
transmission signal X'2(n, k) . The signal separation &
equalization unit 39 then outputs the first transmission signal

X'1(n, k) and the second transmission signal X'2(n, k).
[0165]
«Second Embodiment>>
The following describes a second embodiment of the present
invention with reference to the drawings . Similarly to the first
embodiment, the present embodiment relates to an MIMO
transmission system that includes : a transmitter 10a having two
transmitting antennas 11 and 12; and a receiver 30a having two
receiving antennas 31 and 32 . Yet, SP signals transmitted from
the second transmitting antenna 12 are different from those
described in the first embodiment. In the following description
of the present embodiment, the same components as those employed
in the first embodiment are denoted by the same reference numerals,
and the description thereof is omitted since the corresponding
description given in the first embodiment is applicable.
[0166]
≤SP Signals>
Before the detailed description of the transmitter 10a and
the receiver 30a according to the present embodiment, a
description is given of the description of SP signals transmitted
from the first transmitting antenna 11 and the second
transmitting antenna 12 both of which are included in the
transmitter 10a.
[0167]
The SP signal transmitted from the first transmitting
antenna 11 are arranged in the pattern shown in FIG. 17 described
above. The complex number of each SP signal allocated to a cell

is the same as the complex number of a typical SP signal allocated
to the same cell according to the DVB-T and ISDB-T systems.
[0168]
In contrast, SP signals transmitted from the second
transmitting antenna 12 are arranged in the pattern shown in
FIG. 9. Note that a plus ( + ) sign in FIG. 9 indicates that the
polarity of each SP signal transmitted from the second
transmitting antenna 12 in a cell represented by a plus ( + ) sign
is not inverted with respect to the polarity of an SP signal
transmitted from the first transmitting antenna 11 in a
corresponding cell having the same symbol number and the same
carrier number. On the other hand, a minus (-) sign in FIG.
9 indicates that the polarity of each SP signal transmitted from
the second transmitting antenna 12 in a cell represented by a
minus (-) sign is inverted with respect to the polarity of an
SP signal transmitted from the first transmitting antenna 11
in a corresponding cell having the same symbol number and the
same carrier number.
[0169]
As shown in FIGs. 17 and 9, the SP signals transmitted from
the first transmitting antenna 11 are arranged in the same pattern
as the SP signals transmitted from the second transmitting
antenna 12. In addition, in both the patterns of the SP signal
arrangement, each cell carrying an SP signal within a symbol
having the symbol number n has a carrier number kSP(n) that
satisfies Equation 1 described above.
[0170]
The SP signals transmitted from the second transmitting

antenna 12 in one symbol are alternately inverted andnot inverted
in polarity with respect to corresponding signals transmitted
from the first transmitting antenna 11 in the same symbol. In
the direction in which the symbol number is incremented by 3
and the carrier number is decremented by 3, the polarity of all
the SP signals are inverted with respect to the polarity of
corresponding SP signals transmitted from the first transmitting
antenna 11.
[0171]
The process of inverting and not inverting the polarity
of SP signals transmitted from the first transmitting antenna
11 is equivalent to an arithmetic operation of multiplying
individual SP signals transmitted from the first transmitting
antenna 11, by the complex plane wave expressed by the left side
of Equation 19 shown below. Here, the complex plane wave has
an equτ-phase line parallel to the k axis direction on the k-n
plane, and the cycle in the n axis direction is equal to 8n and
the cycle in the k axis direction is -8k.
[0172]
[Equation 19]

Note that in Equation 19, the right side is obtained by
.rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0173]
Accordingly, the response of each SP signal transmitted

from the second transmitting antenna 12 is said to be shifted
the response of a corresponding SP signal transmitted from the
first transmitting antenna 11, by -Tu/8 in the τ axis direction
and 1/(8Ts) in the fD axis direction on the τ-fD plane.
[0174]
Inviewof the above, the responses of SP signals transmitted
from the first transmitting antenna 11 and the responses of SP
signal transmitted from the second transmitting antenna 12 are
expressed on the same τ-fD plane as shown in FIG. 10. Note that
a black dot in FIG. 10 represents a response of an SP signal
transmitted from the first transmitting antenna 11, whereas a
cross represents a response of an SP signal transmitted from
the second transmitting antenna 12.
[0175]
The receiver 3 0a divides each received SP signal by the
nominal SP signal to obtain a mixed channel response which is
a mixture of a channel response of the channel from the first
transmitting antenna 11 to one of the receiving antennas (the
first receiving antenna 31 or the second receiving antenna 32)
and a channel response of the channel from the second transmitting
antenna 12 to the one of the receiving antennas (the former channel
response is referred as the "channel response related to the
first transmitting antenna 11" and the latter is referred to
as the "channel response related to the second transmitting
antenna 12").
[0176]
However, the channel response related to the first
transmitting antenna 11 has the spreading from the black dots

shown in FIG. 10, in accordance with the impulse response and
Doppler spectrum. Similarly, the channel response related to
the second transmitting antenna 12 has the spreading from the
crosses shown in FIG. 10, in accordance with the impulse response
and Doppler spectrum.
[0177]
FIG. 11 shows a region the τ-fD plane in which the respective
channel responses related to the first and second transmitting
antennas 11 and 12 are interpolated without causing aliasing
distortion and separated from each other without causing
crosstalk therebetween, on condition that the channel responses
of SP signals are interpolated first in the n axis direction
and then in the k axis direction on the k-n plane. In FIG. 11,
a black dot represents a response of an SP signal transmitted
from the first transmitting antenna 11, whereas a cross
represents a response of an SP signal transmitted from the second
transmitting antenna 12 . Inaddition, a rectangular witha solid
line represents the channel response related to the first
transmitting antenna 11, whereas a rectangular with a broken
line represents the response related to the second transmitting
antenna 12.
[0178]
From FIG. 11, it is known that a rectangular region having
a width of Tu/3 in the τ axis direction and a width of l/(8Ts)
in the fD axis direction is what is hereinafter refereed to as
an "interpolatable & separable region" . In the interpolatable
& separable region, the channel response related to the first
transmitting antenna 11 and the channel response related to the

second transmitting antenna 12 are interoperated without causing
aliasing distortion and separated without causing crosstalk
therebetween.
[0179]
From a comparison of FIG. 22 with FIG. 11, the following
is noted. On condition that SP signals transmitted from the
first transmitting antenna 11 are arranged in the pattern shown
in FIG. 17, whereas SP signals transmitted from the second
transmitting antenna 12 are arranged in the pattern shown in
FIG. 9. Then, the width of the interpolatable & separable region
in the x axis direction is Tu/3, which is equal to the width
of the interpolatable region in the x axis direction shown in
FIG. 17. As mentioned above, the interpolatable region shown
in FIG. 17 is a region in which SP signals transmitted from a
single transmitting antenna are interpolated without causing
aliasing distortion. That is, the T axis-direction width of
the region in which correct estimation is ensured for both the
channel responses related to the first and second transmitting
antennas 11 and 12 falls within a range not impairing the tolerance
to multτ-path delay achieved by the insertion of guard intervals .
[0180]
FIG. 12 shows a region of the τ-fD plane in which the channel
response related to the first transmitting antenna 11 and the
channel response related to the second transmitting antenna 12
are interpolated without causing aliasing distortion and
separated from each other without causing crosstalk therebetween,
on condition that the channel response of SP signals are
interpolated only in the k axis direction and not in the n axis

direction on the k-n plane. In FIG. 12, a black dot represents
a response of an SP signal transmitted from the first transmitting
antenna 11, whereas a cross represents a response of an SP signal
transmitted from the second transmitting antenna 12. In
addition, a rectangular with a solid line represents the channel
response related to the first transmitting antenna 11, whereas
a rectangular with a broken line represents the channel response
related to the second transmitting antenna 12.
[0181]
From FIG. 12, it is known that a rectangular region having
a width of Tu/24 in the τ axis direction and a width of 1/Ts
in the fD axis direction is what is hereinafter refereed to as
an "interpolatable & separable region" . In the interpolatable
& separable region, the channel response related to the first
transmitting antenna 11 and the channel response related to the
second transmitting antenna 12 are interoperated without causing
aliasing distortion and separated from each other without causing
crosstalk therebetween.
[0182]
From a comparison of FIG. 23 with FIG. 12, the following
is noted on condition that SP signals transmitted from the first
transmitting antenna 11 are arranged in the pattern shown in
FIG. 17 and that SP signals transmitted from the second
transmitting antenna 12 are arranged in the pattern shown in
FIG. 9. That is, the width of the interpolatable & separable
region in the fD axis direction is 1/Ts, which is equal to the
width of the interpolatable region in the τ axis direction shown
in FIG. 17. As mentioned above, the interpolatable region shown

in FIG. 17 is a region in which the channel response of SP signals
transmitted from a single transmitting antenna is interoperated
without causing aliasing distortion. That is, the fD-axis
direction width of the region in which correct estimation is
ensured for both the channel responses related to the first and
second transmitting antennas 11 and 12 is not impaired at all.
In other words, the ability of following the channel's time
variability is not impaired.
[0183]
Further, FIG. 24 and FIG. 10 are compared.
[0184]
In FIG. 24, the responses appearing at the same Doppler
frequency are a mixture of the responses of SP signals transmitted
from the first transmitting antenna and the responses of SP signal
transmitted from the second transmitting antenna. Therefore,
the impulse responses of the respective channels share the same
region in the τ axis direction. Also, the responses appearing
at the same delay time are a mixture of the responses of SP signals
transmitted from the first transmitting antenna and the responses
of SP signals transmitted from the second transmitting antenna.
Therefore, the Doppler spectrums of the respective channels share
the same region in the fD axis direction.
[0185]
In FIG. 10, in contrast, the responses appearing at the
same Doppler frequency are exclusively of the responses of SP
signals transmitted fromeither of the first transmitting antenna
11 and the second transmitting antenna 12. Therefore, the
impulse responses of the respective channels are allowed to

occupy mutually different regions in the τ axis direction. Also,
the responses appearing at the same delay time are exclusively
of the responses of SP signals transmitted from either of the
first transmitting antenna 11 and the second transmittingantenna
12 . Therefore, the Doppler spectrums of the respective channels
are allowed to occupy mutually different regions in the fD axis
direction.
[0186]
As described above, according to the present embodiment
that uses the two SP signal arrangements shown in FIGs. 17 and
9, the interpolatable & separable region is extended widthwise
in the T or fD axis direction, as compared with the interpolatable
& separable region in the conventional case where the two SP
signal arrangements shown in FIGs. 17 and 20 are used. As
described above, the interpolatable & separable region refers
to a region in which the channel response related to the first
transmitting antenna 11 and the channel response related to the
second transmitting antenna 12 are interpolated without causing
aliasing distort ion and separated from each other without causing
crosstalk therebetween.
[0187]

The following describes the structure of the transmitter
10a with reference to FIG. 13. FIG. 13 is a diagram showing
the structure of the transmitter 10a according to the present
embodiment.
[0188]
The transmitter 10a includes a complex plane wave generation

unit 14a, instead of the complex plane wave generation unit 14
included in the transmitter 10.
[0189]
[Complex Plane Wave Generation Unit 14a]
The complex plane wave generation unit 14a generates a
complex plane wave expressed by Equation 2 0 below and outputs
the generated complex plane wave to the multiplication unit 15.
[0190]
[Equation 20]

In Equation 20, n represents the symbol number and k
represents the carrier number.
[0191]
Note that the multiplication unit 15 according to the
present embodiment multiplies an SP signal received from the
SP generation unit 13 by the complex plane wave received from
the complex plane wave generation unit 14a, rather than by the
complex plane wave received from the complex plane wave
generation unit 14. The multiplication unit 15 then outputs
the result of the multiplication to the cell allocation unit
21 where the received multiplication result is used as an SP
signal to be transmitted from the second transmitting antenna
12. The process of the multiplication performed by the
multiplication unit 15 is equivalent to the process of shifting
the SP signal received from the SP generation unit 13, by -Tu/8

in the T axis direction and by 1/ (8Ts) in the fD axis direction
on the τ-fD plane.
[0192]

The following describes the structure of the receiver 30a
according to the present embodiment, with reference to FIG. 14.
FIG. 14 is a diagram showing the structure of the receiver 30a
according to the present embodiment.
[0193]
The receiver 3 0a includes channel separation & estimation
units 35a and 38a, instead of the channel separation & estimation
units 35 and 38 included in the receiver 30.
[0194]
[Channel Separation & Estimation Unit 35a]
The channel separation & estimation unit 3 5a separates and
estimates the channel response H'11(n, k) of the channel P11
and the channel response H'12(n, k) of the channel P12, with
the use of the first reception signal Y'1(n, k) received from
the Fourier transform unit 34. The channel separation &
estimation unit 35a then outputs the channel responses H'11(n, k) and H'12 (n, k) to the signal separation & equalization unit
39.
[0195]
[Channel Separation & Estimation Unit 38a]
The channel separation & estimation unit 38a separates and
estimates the channel response H'21(n, k) of the channel P21
and the channel response H'22(n, k) of the channel P22, with
the use of the second reception signal Y'2 (n, k) received from

the Fourier transform unit 37. The channel separation &
estimation unit 38a then outputs the channel responses H'21(n,
k) and H'22(n, k) to the signal separation & equalization unit
39.
[0196]
≤Structures of Channel Separation & Estimation Units 35a and
38a>
The following describes the details of the channel
separation & estimation units 35a and 38a shown in FIG. 14, with
reference to FIG. 15. FIG. 15 is a diagram showing the structures
of the channel separation & estimation units 35a and 38a shown
in FIG. 14.
[0197]
[Channel Separation & Estimation Unit 35a]
The channel separation & estimation unit 35a includes a
complex plane wave generation unit 55a instead of the complex
plane wave generation unit 55 included in the channel separation
& estimation unit 35.
[0198]
(Complex Plane Wave Generation Unit 55a)
The complex plane wave generation unit 55a generates a
complex plane wave expressed by Equation 21 shown below and
outputs the generated complex plane wave to the multiplication
unit 56 and the multiplication unit 66 that is included in the
channel separation & estimation unit 38a.
[0199]
[Equation 21]


Note in Equation 21, n represents the symbol number and
k represents the carrier number.
[0200]
Note that the multiplication unit 56 according to the
present embodiment multiplies the division result (the first
mixed channel response) received from the division unit 53 by
the complex plane wave received from the complex plane wave
generation unit 55a, rather than by the complex plane wave
received from the complex plane wave generation unit 55. The
multiplication unit 56 then outputs the result of the
multiplication (the first shifted and mixed channel response)
to the interpolation unit 57. Note that the multiplication
performed by the multiplication unit 56 is equivalent to the
process of shifting the first mixed channel response by Tu/8
in the τ axis direction and by -1/ (8Ts) in the fD axis direction
on the τ-fD plane.
[0201]
[Channel Separation & Estimation unit 38a]
The channel separation & estimation unit 38a is of the same
structure as the channel separation & estimation unit 38. Note
that the multiplication unit 66 according to the present
embodiment multiplies the division result (the second mixed
channel response) received from the division unit 63 by the
complex plane wave received from the complex plane wave
generation unit 55a, rather than by the complex plane wave

received from the complex plane wave generation unit 55. The
channel separation & estimation unit 38a then outputs the result
of the multiplication unit (the second shifted and mixed channel
response) to the interpolation unit 67. Note that the
multiplication performed by the multiplication unit 66 is
equivalent to the process of shifting the second mixed channel
response by Tu/8 in the τ axis direction and -1/(8Ts) in the
fD axis direction on the τ-fD plane.
[0202]
<>
The present invention is not limited to the specific
embodiments described above. Various modifications including
the following still falls within the scope of the present
invention.
[0203]
(1) The above embodiments are described using the term "SP
signal", on the assumption that the DVB-T or ISDB-T system is
employed. It should be noted, however, that the present
invention is applicable to various other cases where a pilot
signal having the amplitude and phase known to receivers are
used.
[0204]
(2) The first embodiment described above may be modified,
so that the complex plane wave generation unit 14 generates a
complex plane wave having a phase term calculated by multiplying
the phase term of Equation 16 by -1, and that the complex plane
wave generation unit 55 generates a complex plane wave having

a phase term calculated by multiplying the phase term of Equation
18 by -1.
[0205]
The second embodiment described above may be modified, so
that the complex plane wave generation unit 14a generates a
complex plane wave having a phase term calculated by multiplying
the phase term of Equation 20 by -1, and that the complex plane
wave generation unit 55a generates a complex plane wave having
a phase term calculated by multiplying the phase term of Equation
21 by -1.
[0206]
(3) The above embodiments are described on the assumption
that the DVB-T or ISDB-T system is employed, so that the interval
between SP signals within the same symbol is described as 12
carriers and the interval between SP signals within the same
carrier is described as 4 symbols. Then, the carrier number
kSP(n) of an SP signal satisfies Equation 1. Yet, the present
invention is applicable to a case where the arrangement pattern
of pilot signals (i.e., signals whose amplitude and phase are
known to receivers) on the carrier-symbol plane is generalized
as follows.
[0207]
Let n denote the symbol number, k denote the carrier number,
Ak denote the interval between pilot signals in the same symbol,
An denote the interval between pilot signals in the same carrier,
and p denote an integer greater than or equal to 0. Then, the
carrier number kP(n) of a carrier transmitting a pilot signal
satisfies Equation 22 shown below.

[0208]
[Equation 22]

In this case, the first or second embodiment may be modified
in the following manner.
[0209]
(3-A) FIG. 16 shows the responses of SP signals transmitted
from the first transmitting antenna in a generalized pattern
of the SP signal arrangements satisfying Equation 22. In FIG.
16, the distance between points A and B is equal to Tu/Ak in
the τ axis direction and to 1/(Ts∆n) in the fD axis direction.
Each pilot signal to be transmitted from the second transmitting
antenna is so adjusted to coincide with a midpoint between the
points A and B.
[0210]
Accordingly, pilot signals to be transmitted from the second
transmitting antenna are generated by shifting the response of
a corresponding pilot signal transmitted from the first
transmitting antenna, by Tu/(2Ak) in the τ axis direction and
1/(2TsAn) in the fD axis direction.
[0211]
In view of the above, the pilot signal generation unit
provided within the transmitter is configured to generate
reference pilot signals (i.e., signals whose amplitude and phase
are known to receivers, and the same holds true for the following
description) and use the thus generated reference pilot signals

as pilot signals to be transmitted from the first transmitting
antenna. In view of the above, in addition, the complex plane
wave generation unit provided within the transmitter is
configured to generate a complex plane wave expressed by the
left side of Equation 23 shown below. The multiplication unit
is configured to multiply each reference pilot signal by the
thus generated complex plane wave and use the result of the
multiplication as a pilot signal to be transmitted from the second
transmitting antenna.
[0212]
[Equation 23]

Note that in Equation 23, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0213]
The receiver divides each pilot signal received with the
receiving antenna (the first or e second receiving antenna),
by the nominal pilot signal (a reference pilot signal generated
by the transmitter, the same holds true for the following
description) . As a division result, the receiver obtains what
is hereinafter referred to as a "mixed channel response" . The
mixed channel response is a mixture of the channel response of
a channel from the first transmitting antenna to that receiving
antenna (the channel response related to the first transmitting
antenna) and the channel response of a channel from the second

transmitting antenna to that receiving antenna (the channel
response related to the second transmitting antenna). The
interpolation unit provided for the first transmitting antenna
estimates, with the use of a low-pass filter, the channel response
at every cell and related to the first transmitting antenna from
the mixed channel response. The complex plane wave generation
unit provided within the receiver generates a complex plane wave
expressed by the left side of Equation 24 shown below. The
multiplication unit then multiples the mixed channel response
by the thus generated complex plane wave. On the other hand,
the interpolation unit provided for the second transmitting
antenna estimates, with the use of a low-pass filter, the channel
response related to the second transmitting antenna at every
cell from the result of the multiplication by the multiplication
unit.
[0214]
[Equation 24]

Note that in Equation 24, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0215]
Note that the complex plane wave generation unit included
in the transmitter may be configured to generate a complex plane
wave having the phase term calculated by multiplying the phase
term of Equation 23 by -1. Then, the complex plane wave

generation unit provided within the receiver may be configured
to generate a complex plane wave having the phase term calculated
by multiplying the phase term of Equation 24 by -1.
[0216]
(3-B) With reference to FIG. 16 showing the response of
pilot signals transmitted from the first transmitting antenna,
the distance between points A and C is equal to 3Tu/Ak in the
τ axis direction and to 1/ (Ts∆n) in the fD axis direction. Each
pilot signal to be transmitted from the second transmitting
antenna is so adjusted to coincide with a midpoint between the
points A and C.
[0217]
Such pilot signals to be transmitted from the second
transmitting antenna are generated by shifting the response of
a corresponding pilot signal transmitted from the first
transmitting antenna, by -3Tu/ (2Ak) in the T axis direction and
by 1/(2TsAn)in the fD axis direction.
[0218]
In view of the above, the pilot signal generation unit
provided within the transmitter is configured to generate
reference pilot signals and use the generated reference pilot
signals as pilot signals to be transmitted from the first
transmitting antenna. In view of the above, in addition, the
complex plane wave generation unit provided within the
transmitter is configured to generate a complex plane wave
expressed by the left side of Equation 25 shown below. The
multiplication unit is configured to multiply each reference
pilot signal by the thus generated complex plane wave and use

the result of the multiplication as a pilot signal to be
transmitted from the second transmitting antenna.
[0219]
[Equation 25]

Note that in Equation 25, the right side is obtained by-
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0220]
The receiver divides each pilot signal received with the
receiving antenna (the first or second receiving antenna), by
a nominal pilot signal. As a result of the division, the
receiver obtains what is hereinafter referred to as a "mixed
channel response" . The mixed channel response is a mixture of
the channel response of the channel from the first transmitting
antenna to that receiving antenna (the channel response related
to the first transmitting antenna) and the channel response of
the channel from the second transmitting antenna to that
receiving antenna (the channel response related to the second
transmitting antenna) . The interpolation unit provided for the
first transmitting antenna estimates, with the use of a low-pass
filter, the channel response at every cell and related to the
first transmitting antenna from the mixed channel response. The
complex plane wave generation unit provided within the receiver
generates a complex plane wave expressed by the left side of
Equation 26, and the multiplication unit multiplies each mixed

channel response by the complex plane wave. The interpolation
unit provided for the second transmitting antenna estimates,
with the use of a low-pass filter, the channel response at every
cell and related to the second transmitting antenna from the
result of the multiplication calculated by the multiplication
unit.
[0221]
[Equation 26]

Note that in Equation 26, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0222]
Note that the complex plane wave generation unit provided
within the transmitter may generate a complex plane wave having
a phase term calculated by multiplying the phase term of Equation
25 by -1, and that the complex plane wave generation unit provided
within the receiver may generate a complex plane wave having
a phase term calculated by multiplying the phase term of Equation
26 by -1.
[0223]
(4) According to the above embodiments, transmission
signals are transmitted from the first transmitting antenna 11
and the second transmitting antenna 12. However, the
transmitter may have M transmitting antennas (M is an integer
greater than or equal to 2) . Here, each of a set of pilot signals

(signals whose amplitude and phase are known to receivers)
transmitted from the mth transmitting antenna (m is an integer
satisfying 1 ≤ m ≤ M) is referred to as an mth pilot signal and
arranged in the generalized pattern expressed by Equation 22
shown above.
[0224]
In this case, the first or second embodiment described above
may be modified in the following manner, for example.
[0225]
(4-A) With reference to FIG. 16 showing the responses of
SP signals transmitted from the first transmitting antenna, the
distance between the points A and B is equal to Tu/Ak in the
τ axis direction and to 1/ (Ts∆n) in the fD axis direction. Each
of themth pilot signals, which are pilot signals to be transmitted
from the mth transmitting antenna (where 2 ≤ m ≤ M) , is so adjusted
that the response of the pilot signal coincides with the (m-1)th
point from the point A, out of (M-1) points determined by dividing
the distance between the points A and B into M segments at even
intervals.
[0226]
Each of the mth pilot signals, which are pilot signals to
be transmitted from the mth transmitting antenna (where 2 ≤ m
≤ M) , is generated by shifting the response of a corresponding
first pilot signal transmitted from the first transmitting
antenna, by (m-1) Tu/(MAk) in the τ axis direction and
(m-1)/(MTsAn) in the fD axis direction.
[0227]
In view of the above, the pilot signal generation unit

provided within the transmitter is configured to generate
reference pilot signals and use the thus generated reference
pilot signals as first pilot signals to be transmitted from the
first transmitting antenna. For each mth transmitting antenna
were 2 ≤ m ≤ M, the complex plane wave generation unit provided
within the transmitter is configured in view of the above to
generate a complex plane wave expressed by the left side of
Equation 27 . The multiplication unit is configured to multiply-
each reference pilot signal by the thus generated complex plane
wave and use the result of the multiplication as an mth pilot
signal to be transmitted from the mth transmitting antenna.
[0228]
[Equation 27]

Note that in Equation 27, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0229]
The receiver divides each pilot signal received with the
receiving antenna by the nominal pilot signal to obtain what
is hereinafter referred to as a "mixed channel response" . The
mixed channel response is a mixture of the multiple channels
from the multiple transmit ting antennas to the receiving antenna.
The interpolation unit provided for the first transmitting
antenna estimates, with the use of a low-pass filter, the channel
response at every cell and related to the first transmitting

antenna, from the mixed channel response. For estimation of
the response related to each mth transmitting antenna were 2
≤ m ≤ M, the complex plane wave generation unit provided within
the receiver generates a complex plane wave expressed by the
left side of Equation 28 shown below. The multiplication unit
multiplies each reference pilot signal by the thus generated
complex plane wave. The interpolation unit estimates, with the
use of a low-pass filter, the channel response at every cell
and related to the mth transmitting antenna from the result of
the multiplication computed by the multiplication unit.
[0230]
[Equation 28]

Note that in Equation 28, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0231]
The complex plane wave generation unit provided within the
transmitter may be modified to generate a complex plane wave
having the phase term calculated by multiplying the phase term
of Equation 27 by -1. Then, the complex plane wave generation
unit provided within the receiver may be modified to generate
a complex plane wave having the phase term calculated by
multiplying the phase term of Equation 28 by -1.
[0232]
(4-B) With reference to FIG. 16 showing the responses of

SP signals transmitted from the first transmitting antenna, the
distance between the points A and C is equal to 3Tu/∆k in the
x axis direction and to 1/ (Ts∆n) in the fD axis direction. Each
of a set of mth pilot signals to be transmitted from the mth
transmitting antenna (where 2 ≤ m ≤ M) is so adjusted that the
response of the pilot signal coincides with the (m-1)th point
from the point A, out of (m-1) points determined by dividing
the distance between the points A and C into M segments at even
intervals.
[0233]
Each of the mth pilot signals, which are pilot signals to
be transmitted from the mth transmitting antenna (where 2 ≤ m
≤ M) , is generated by shifting the response of a corresponding
first pilot signal transmitted from the first transmitting
antenna, by -3(m-1)Tu/(MAk) in the x axis direction and
(m-1)/(MTs∆n) in the fD axis direction.
[0234]
In viewof the above, thepilot signal generation unit within
the transmitter generates reference pilot signals and use the
thus generated reference pilot signals as first pilot signals
to be transmitted from the first transmitting antenna. In
addition, the complex plane wave generation unit provided within
the transmitter is configured in view of the above to generate
a complex plane wave expressed by the left side of Equation 29
for each mth transmitting antenna, were 2 ≤ m ≤ M. The
multiplication unit multiplies each reference pilot signal by
the thus generated complex plane wave and uses the result of
the multiplication as an mth pilot signal to be transmitted from

the mth transmitting antenna.
[0235]
[Equation 29]

Note that in Equation 29, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0236]
The receiver divides each pilot signal received with the
receiving antenna by the nominal pilot signal to obtain what
is hereinafter referred to as a "mixed channel response" . The
mixed channel response is a mixture of the channel responses
of the multiple channels from the multiple transmitting antennas
to the receiving antenna. The interpolation unit provided for
the first transmitting antenna estimates, with the use of a
low-pass filter, the channel response at every cell and related
to the first transmitting antenna, from the mixed channel
response. For estimation of each mth transmitting antenna were
2 ≤ m ≤ M, the complex plane wave generation unit provided within
the receiver generates a complex plane wave expressed by the
left side of Equation 30 shown below. The multiplication unit
multiplies each reference pilot signal by the thus generated
complex plane wave. The interpolation unit estimates, with the
use of a low-pass filter, the channel response at every cell
and related to the mth transmitting antenna from the result of
the multiplication.

[0237]
[Equation 30]

Note that in Equation 30, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0238]
The complex plane wave generation unit provided within the
transmitter may be modified to generate a complex plane wave
having the phase term calculated by multiplying the phase term
of Equation 29 by -1. Then, the complex plane wave generation
unit provided within the receiver may be modified to generate
a complex plane wave having the phase term calculated by
multiplying the phase term of Equation 30 by -1.
[0239]
In addition, the modification described in (4-B) above
requires that -3(m-1) is not equal to an integral multiple of
M, where 2 ≤ m ≤ M.
[0240]
(5) With reference to FIG. 16, the following describes a
further generalization of the SP signal transmission methods
according to the embodiments described above. In FIG. 16, a
black dot represents a response of a first pilot signal
transmitted from the first transmitting antenna, whereas a cross
represents the response of a second pilot signal transmitted
from the second transmitting antenna. Here, let M denote the

number of transmitting antennas (where M is an integer greater
than or equal to 2) . Then, a set of mth pilot signals transmitted
from themth transmitting antenna (where m is an integer satisfying
1 ≤m≤M) are arranged in the generalized pat tern given by Equation
22 shown above.
[0241]
In FIG. 16, solid straight lines extend in parallel to the
Τ or fD axis to pass through the responses of first pilot signals.
Similarly, doted straight lines extend in parallel to the τ or
fD axis to pass through the responses of second pilot signals.
[0242]
The gist of the method for transmitting pilot signals from
multiple transmitting antennas is to ensure that the responses
of pilot signals transmitted from one transmitting antenna do
not appear on the same lattice pattern with the responses of
pilot signals transmitted from any other transmitting antenna.
[0243]
The above signal arrangement is realized in the following
manner. First, the distance between each adjacent lines in the
lattice pattern of the responses of first pilot signals
transmitted from the first transmitting antenna is divided by
M to define M different lattice patterns. Then, sets of pilot
signals transmitted from the respective transmitting antennas
are so adjusted that their responses appear onmutually different
lattice patterns.
[0244]
In view of the above, the pilot signal generation unit
provided within the transmitter is configured to generate

reference pilot signals and use the thus generated reference
pilot signals as first pilot signals to be transmitted from the
first transmitting antenna. Also in view of the above, in
addition, the complex plane wave generation unit provided within
the transmitter is configured to generate a complex plane wave
expressed by the left side of Equation 31 for each mth transmitting
antenna (where 2 ≤ m ≤ M) . Then, the multiplication unit
multiplies each reference pilot signal by the thus generated
complex plane wave and the result of the multiplication is used
as a mth pilot signal to be transmitted from the mth transmitting
antenna.
[0245]
[Equation 31]

Note that where 2 ≤ m ≤ M, ns and ks are nonzero integers
such that neither (m-1)ns nor (m-1)ks is an integral multiple
of M.
[0246]
In Equation 31 shown above, the right side is obtained by
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0247]
In addition, Equation 31 above is equivalent to a process
of shifting first pilot signals to a different lattice pattern
defined by the M-division of the original lattice pattern. In
addition, the condition required by Equation 31 that ns and ks

are nonzero integers such that neither (m-1)ns nor (m-1)ks is
an integral multiple of M is to ensure that any of the lattice
patterns of the pilot signal responses coincide with another.
[0248]
The receiver divides each pilot signal received with the
receiving antenna by a nominal pilot signal to obtain what is
hereinafter referred to as a "mixed channel response" . The mixed
channel response is a mixture of the channel responses of the
multiple channels from the multiple transmitting antennas to
the receiving antenna. The interpolation unit provided for the
first transmitting antenna estimates, with the use of a low-pass
filter, the channel response at every cell and related to the
first transmitting antenna, from the mixed channel response.
For estimation of the response related to each mth transmitting
antenna were 2 ≤ m ≤ M, the complex plane wave generation unit
provided within the receiver generates a complex plane wave
expressed by the left side of Equation 32 shown below. The
multiplication unit multiplies the mixed channel response by
the thus generated complex plane wave. The interpolation unit
then estimates, with the use of a low-pass filter, the channel
responses at every cell andrelated to the mth transmitting antenna
from the result of the multiplication computed by the
multiplication unit.
[0249]
[Equation 32]


Note that in Equation 32, the right side is obtained by-
rewriting the left side using the relation n = (1/Ts)t and k
= Tuf.
[0250]
(6) In the embodiments and modifications described above,
theprocess of shifting the phase of a signal (SPsignal, reception
SP signal, or the result obtained by the division by a nominal
SP signal) is carried out by generating a complex plane wave
followed by multiplication of the signal by the thus generated
complex plane wave. It is noted, however, this phase shift
process is merely one example and without limitation. The phase
shift may be carried out in the following manner, for example.
[0251]
(6-A) Here, let M denote the number of transmitting antennas
(M is an integer greater than or equal to 2) and a set of mth
pilot signals transmitted from the mth transmitting antenna
(where m is an integer satisfying 1 ≤ m ≤ M) are arranged in
the generalized pattern given by Equation 22 shown above.
[0252]
The pilot signal generation unit provided within the
transmitter generates reference pilot signals and use the thus
generated reference pilot signals as first pilot signals to be
transmitted from the first transmitting antenna. For each mth
transmitting antenna where 2 ≤ m ≤ M (M is an integer greater
than or equal to 2) , the phase generation unit provided within
the transmitter generates the phase of any of the complex plane
waves generated by the complex plane wave generation unit also
provided within the transmitter mentioned above. Then, a phase

rotation unit also provided within the transmitter rotates the
phase of each reference pilot signal by the amount corresponding
to the phase generated by the phase generation unit. The
reference pilot signals after the phase rotation are used as
mth pilot signals, which are a set of pilot signals to be
transmitted from the mth transmitting antenna. Note that the
phase rotation may be performed by employing a known algorithm
such as CORDIC.
[0253]
The receiver divides each pilot signal received with the
receiving antenna by a nominal pilot signal to obtain what is
hereinafter referred to as a "mixed channel response" . The mixed
channel response is a mixture of the channel responses of the
multiple channels from the multiple transmitting antennas to
the receiving antenna. The interpolation unit provided for the
first transmitting antenna estimates, with the use of a low-pass
filter, the channel response at every cell and related to the
first transmitting antenna. For estimation each mth
transmitting antenna where 2 ≤ m ≤ M (M is an integer greater
than or equal to 2) , the phase generation unit provided within
the receiver generates the phase of any of the complex plane
waves generated by the complex plane wave generation unit that
is also provided within the receiver and correspond to the phase
generation unit provided within the transmitter. Then, the
phase rotation unit also provided within the receiver rotates
the phase of each reference pilot signal by the amount
corresponding to the phase generated by the phase generation
unit. The interpolation unit estimates, with the use of a

low-pass filter, the channel response at every cell and related
to the mth transmitting antenna from the mixed channel response
after the phase rotation. Note that the phase rotation may be
performed by employing a known algorithm such as CORDIC.
[0254]
Regarding the transmitter, the complex plane wave
generation unit and the multiplication unit are replaced by the
phase generation unit and the phase rotation unit. Regarding
the receiver, the complex plane wave generation unit and the
multiplication unit are replaced by the phase generation unit
and the phase rotation unit.
[0255]
(6-B) The pilot signal generation unit provided within the
transmitter generates reference pilot signals and uses the thus
generated reference pilot signals as pilot signals to be
transmitted from the first transmitting antenna. The
transmitter is provided with a polarity inversion unit instead
of the complex plane wave generation unit and the multiplication
unit. The polarity inversion unit inverts and does not invert
the polarity of every other reference pilot signal within the
same symbol, thereby generating pilot signals to be transmitted
from the second transmitting antenna.
[0256]
(7) In the above description of the embodiments, SP signals
transmitted from a transmitting antenna other than the first
transmitting antenna are generated by the multiplication, phase
rotation, andpolarityreversal of a complexplane wave. However,
this is merely one example and without limitation. SP signal

transmitted from a transmitting antenna other than the first
transmitting antenna may be generated in the following manner,
for example.
[0257]
Let M denote the number of transmitting antennas (where
M is an integer greater than or equal to 2), and a set of mth
pilot signals transmitted from the mth transmitting antenna
(where m is an integer satisfying 1 ≤ m ≤ M) are arranged in
the generalized pattern given by Equation 22 shown above.
[0258]
It is sufficient to provide a means for generating a
plurality of mth pilot signals, which are a set of pilot signals
to be transmitted from the mth transmitting antenna (where m
is an integer satisfying 1 ≤ m ≤ M) , in a manner that the phase
difference between each mth pilot signal and the reference pilot
signal is equal to the value given by Equation 33 shown below.
[0259]
[Equation 33]

Mote that it is required that when 2 ≤ m ≤ M, ns and ks
are nonzero integers such that neither (m-1)ns nor (m-1)ks is
equal to an integral multiple of M.
[0260]
(8) In the above description, the transmitter has M
transmitting antennas (where M is an integer greater than or
equal to 2). Here, M is the maximum number of transmitting

antennas that the transmitter may use for transmission. In
practice, however, the number of transmitting antennas actually
used may be equal to or smaller than the number M.
[0261]
(9) According to the above-described embodiments, the
receiver conducts the processing sequentially by the division
units 53 and 63, the multiplication units 56 and 66, the low-pass
filters included in the interpolation units 57 and 64, in the
stated order. It should be noted, however, that the processing
order is not limited to such. For example, the processing by
the respective units may be conducted in the order of the division
units 53 and 63 , high-pass filters which are used as alternatives
to low-pass filters, and the multiplication units 56 and 66.
[0262]
(10) It is possible to provide the transmitter with a
mechanism for selectively switching between the SP signal
transmission method according to the above described embodiments
and a conventional SP signal transmission method.
[0263]
The following are examples in which the SP signal
transmission method according to the above embodiments is to
be selected and executed. One is the case where the guard
interval duration that is longer than the useful symbol duration
is selected (for example, Tu/4) and the impulse response of the
channel has the delay to the same extent as the guard interval
duration. Another is the case where the delay of the impulse
response is relatively short and the ability to follow the
channel's time variability is to be improved.

[0264]
On the other hand, the following are examples in which a
conventional SP signal transmission method is to be selected
and executed. One is the case where the guard interval duration
shorter than the useful symbol duration is selected. Another
is the case where the delay spread of the impulse response and
the channel's time variability are both relatively small.
[0265]
Note that the transmitter may be modified to transmit
control information, which is TPC in the case of the DVB-T system
and TMCC in the case of the ISDB-T system, that includes
information indicating the type of the complex plane wave. With
this modification, the receiver is able to select, based on the
control information, a complex plane wave corresponding to the
complex plane wave used at the transmitter side, so that the
subsequent processing is appropriately carried out.
[0266]
In the case of the communications between one transmitter
and one receiver, the receiver may observe the impulse response
and the time variability of the channel to judge which
transmission method is suitable and pass the judgment result
to the transmitter. The transmitter may switch the SP signal
transmission method to a different method according to the
judgment result received from the receiver.
[0267]
(11) In the above embodiments, the MIMO transmission system
is described as an example. However, the present invention is
applicable to a MIMO transmission system having two or more

transmitting antennas and two or more receiving antennas as well
as to MISO (Multiple Input Single Output) transmission system.
[0268]
In addition, the present invention has been described above
by way of an example in which a MIMO transmission system is used
as a spatial multiplexing transmission system. Yet, the present
invention is applicable to a transmission diversity system which
employs a coding scheme, such as STC (Space Time Coding) , STBC
(Space Time Block Coding), and SFBC (Space Frequency Block
Coding).
[0269]
(12) The transmitters and receivers consistent with the
embodiments above may each be realized as an LSI (Large Scale
Integration), which is a type of integrated circuits. The
individual circuits may be implemented on separate chips, or
all or part of the circuits may be implemented on a single chip.
[0270]
Although LSI is specifically mentioned herein, the same
may also be referred to an IC (integrated circuit), a system
LSI, a super LSI, or ultra LSI, depending on the packaging density.
[0271]
The scheme employed for the circuit integration is not
limited to LSI, and the integrated circuit may be implanted by
a dedicated circuit or a general purpose processor. It is also
possible to use an FPGA (Field Programmable Gate Array) , which
allows post manufacture programming of the LSI, or to use a
reconfigurable processor, which allows reconfiguration of the
connection between circuit cells within the LSI or setting of

the circuit cells.
[0272]
Furthermore, if the advance in the field of semiconductor
technology or in another technology derived therefrom introduces
a new integration technology that replaces the LSI, the new
technology may be used to integrate the functional blocks . For
example, the application of biotechnology is one possibility.
[Industrial Applicability]
[0273]
The present invention is applicable to a digital
broadcasting that transmits the transmission parameter
information. The present invention is also to digital
communications which involves the use of mobile telephones,
wireless LAN, power line communications, xDSL, and so on.

CLAIMS
1. A transmitter having first to Mth transmitting antennas
(where M is an integer equal to or greater than 2) and for
transmitting an OFDM signal obtained by modulating a plurality
of carriers per symbol duration,
the OFDM signal containing pilot signals scattered on a
carrier-symbol plane,
on the carrier-symbol plane, k denoting a carrier number,
n denoting a symbol number, Ak denoting an interval between pilot
signals in a same symbol, An denoting an interval between pilot
signals in a same carrier, and p denoting an integer greater
than or equal to 0,
ns and ks each denoting a nonzero integer, and m denoting
an integer satisfying 1 ≤ m ≤ M,
when 2 ≤ m ≤ M, neither (m-1)ns nor (m-1)ks being equal
to an integral multiple of M,
the carrier number kP (n) of a carrier that transmits a pilot
signal in a symbol with the symbol number n satisfying Equation
1,
[Equation 1]

the transmitter comprising:
a generating unit operable to generate a plurality of pilot
signals as mth pilot signals for an mth antenna (where 1 ≤ m ≤
M), such that a phase difference between a phase of each mth
pilot signal and a phase of a reference pilot signal is equal

to a value given by Equation 2,
[Equation 2]

a transmitter operable to transmit, from the mth
transmitting antenna, an OFDM signal containing the mth pilot
signals generated by the generating unit.
2. The transmitter according to Claim 1, wherein
the generating unit includes:
a reference signal generating unit operable to
generate a plurality of first pilot signals, each first pilot
signal being the reference pilot signal; and
a multiplication unit operable to generate the mth
pilot signals where 2 ≤m ≤M, eachmth pilot signal being generated
by multiplying the reference pilot signal by a complex plane
wave expressed by Equation 3 on the carrier-symbol plane,
[Equation 3]

3. The transmitter according to Claim 1, wherein
the generating unit includes:
a reference signal generating unit operable to
generate a plurality of first pilot signals, each first pilot
signal being the reference signal; and
a phase rotation unit operable to generate the mth
pilot signals where 2 ≤m≤M, eachmth pilot signal being generated

by rotating the phase of the reference pilot signal by the value
given by Equation 2.
4. The transmitter according to Claim 1, wherein
M is equal to 2, and
the generating unit includes:
a reference signal generating unit operable to
generate a plurality of first pilot signals, each first pilot
signal being the reference signal; and
a polarity inversion unit operable to generate a
plurality of second pilot signals such that polarities of the
second pilot signals are each alternately inverted and not
inverted in a carrier direction with respect to a polarity of
a corresponding reference signal in a same symbol.
5. The transmitter according to Claim 1, wherein
M is equal to 2,
An is equal to 4,
Ak is equal to 12,
ns is equal to 1, and
ks is equal to 1.
6. The transmitter according to Claim 1, wherein
M is equal to 2,
An is equal to 4,
Ak is equal to 12,
ns is equal to 1, and
ks is equal to -3.

7 . A receiver for receiving an OFDM signal transmitted from
a transmitter having a plurality of first to Mth transmitting
antennas (where M is an integer greater than or equal to 2),
the OFDM signal being obtained by modulating a plurality of
carriers per symbol duration,
the OFDM signal containing a plurality of pilot signals
scattered on a carrier-symbol plane,
on the carrier-symbol plane, k denoting a carrier number,
n denoting a symbol number, ∆k denoting an interval between pilot
signals in a same symbol, ∆n denoting an interval between pilot
signals in a same carrier, and p denoting an integer greater
than or equal to 0,
ns and ks each denoting a nonzero integer, and m denoting
an integer satisfying 1 ≤ m ≤ M,
when 2 ≤ m ≤ M, neither (m-1)ns nor (m-1)ks being equal
to an integral multiple of M,
the carrier number kP (n) of a carrier that transmits a pilot
signal in a symbol with the symbol number n satisfying Equation
4,
[Equation 4]

a plurality of pilot signals transmitted from an mth one
of the transmitting antennas (where m is an integer satisfying
1 ≤ m ≤ M) being mth pilot signals, such that a phase difference
between a phase of each mth pilot signal and a phase of a reference
pilot signal is equal to a value given by Equation 5,

[Equation 5]

the receiver comprising:
a receiving antenna with which the OFDM signal from the
transmitter is received;
a response estimation unit operable to estimate a channel
response of each of first to Mth channels respectively from the
first to Mth antennas to the receiving antenna, the estimation
being carried out based on Equation 5 and pilot signals contained
in the OFDM signal received with the receiving antenna; and
a signal estimation unit operable to estimate first to Mth
transmission signals based on the received OFDM signal and the
estimated channel responses of the first to Mth transmission
channels, the first to Mth transmission signals corresponding
to first to Mth OFDM signals transmitted respectively from the
first to Mth transmitting antennas.
8. The receiver according to Claim 7, wherein
the response estimation unit is operable to extract pilot
signals from the OFDM signal received with the receiving antenna,
divide each extractedpilot signal by the reference pilot signal,
and estimate the channel response of the first channel based
on a result of each division, and
further operable to estimate the channel response of mth
channel where 2 ≤ m ≤ M, based on the result of each division
and Equation 5.

9. An OFDM transmission method for transmitting an OFDM
signal from a transmitter having first to Mth transmitting
antennas (where M is an integer greater than or equal to 2),
the OFDM signal being obtained by modulating a plurality of
carries per symbol duration,
the OFDM signal containing pilot signals scattered on a
carrier-symbol plane,
on the carrier-symbol plane, k denoting a carrier number,
n denoting a symbol number, ∆k denoting an interval between pilot
signals in a same symbol, ∆n denoting an interval between pilot
signals in a same carrier, and p denoting an integer greater
than or equal to 0,
ns and ks each denoting a nonzero integer, and m denoting
an integer satisfying 1 ≤ m ≤ M,
when 2 ≤ m ≤ M, neither (m-1)ns nor (m-1)ks being equal
to an integral multiple of M,
the carrier number kP (n) of a carrier that transmits a pilot
signal in a symbol with the symbol number n satisfying Equation
6,
[Equation 6]

the OFDM transmission method comprising the steps of:
generating a plurality of pilot signals as mth pilot signals
for an mth antenna (where 1 ≤ m ≤ M) , such that a phase difference
between a phase of each mth pilot signal and a phase of a reference
pilot signal is equal to a value given by Equation 7,
[Equation 7]


transmitting, from the mth transmitting antenna, an OFDM
signal containing the mth pilot signals generated in the
generating step.

SP signals to be transmitted from a first transmitting
antenna 11 are arranged in the same pattern as SP signals to
be transmitted from a second transmitting antenna 12 . The SP
signals to be transmitted from the second transmitting antenna
12 in one symbol are generated, such that the polarity of the
SP signals are alternately inverted andnon-inverted with respect
to the SP signals to be transmitted from the first transmitting
antenna 11 in the same symbol. Thus, in the direction that the
symbol number is incremented by 1 and the carrier number is
incremented by 3, the polarity of SP signals transmitted from
the second transmitting antenna 12 are all inverted or
non-inverted with respect to the polarity of corresponding SP
signals transmitted from the first transmitting antenna 11.