A SMALL CELL BASE STATION COMPRISING MULTIPLE ANTENNAS,
AND A METHOD OF CONTROLLING RECEPTION PATTERN BY
SELECTING A SUBSET OF THE ANTENNAS FOR USE
Field of the Invention
The present invention relates to telecommunications, in particular to wireless
telecommunications .
Description of the Related Art
Wireless telecommunications systems are well-known. Many such systems are
cellular, in that radio coverage is provided by a bundle of radio coverage areas known
as cells. A base station that provides radio coverage is located in each cell. Traditional
base stations provide coverage in relatively large geographic areas and the
corresponding cells are often referred to as macrocells.
It is possible to establish smaller sized cells within a macrocell. Cells that are
smaller than macrocells are sometimes referred to as small cells, microcells, picocells,
or femtocells, but we use the terms small cells and femtocells interchangeably and
generically for cells that are smaller than macrocells. One way to establish a femtocell
is to provide a femtocell base station that operates within a relatively limited range
within the coverage area of a macrocell. One example of use of a femtocell base
station is to provide wireless communication coverage within a building.
The femtocell base station is of a relatively low transmit power and hence each
femtocell is of a small coverage area compared to a macrocell. A typical coverage
range is tens of metres.
Femtocell base stations have auto-configuring properties so as to support plugand
play deployment by users, for example in which femto base stations may integrate
themselves into an existing macrocell network so as to connect to the core network of
the macrocell network.
One known type of Femtocell base station uses a broadband Internet Protocol
connection as "backhaul", namely for connecting to the core network. One type of
broadband Internet Protocol connection is a Digital Subscriber Line (DSL). The DSL
connects a DSL transmitter-receiver ("transceiver") of the femtocell base station to the
core network. The DSL allows voice calls and other services provided via the
femtocell base station to be supported. The femtocell base station also includes a radio
frequency (RF) transceiver connected to an antenna for radio communications. An
alternative to such a wired broadband backhaul is to have a wireless backhaul.
Femtocell base stations are sometimes referred to as femtos.
The small coverage range of small cells means that the radio spectrum can be
used much more efficiently than is achievable using macrocells alone, and as the radio
link between small cell base station and a user terminal is small, good quality links are
possible giving high data rates. However, radio interactions between femtos and
macrocells need to be managed. Specifically, radio interference is known to be an
issue to be addressed in joint deployments of femtos within macrocellular networks.
As regards the use of radio spectrum, small cells and macrocells can be
deployed in basically two ways. One is to use dedicated channels so that small cells
use different frequency bands to the macrocells, and hence there is no interference;
and the other is co-channel use where the same frequency bands are used by both
small cells and macrocells. Despite the interference caused, co-channel operation is
desirable as it provides greater use of bandwidth, but interference needs to be
minimised. In addition to interference between femtos and macrocells, the likely
future dense deployment of femtos will require management of interference between
femtos (so-called "intra femtocell interference") even when the macrocells and femtos
use different frequency bands.
Interference may be considered from the perspective of macrocell base stations
or femtocell base stations, and in the downlink direction from a base station or in the
uplink direction to a base station. However, for femtos connected to user terminals,
transmit power is very low in both uplink and downlink directions, particularly as
interference to macrocells by femto users is required to be low. In consequence,
interference caused to a femto by a macrocell base station is normally more significant
than the other way round.
Known femtos use static antenna systems with adjustable output power, where
the antennas are simple dipoles or printed-circuit-board (PCB) antennas. Such
antennas have low gain and basically provide omni-directional fixed coverage
patterns.
Summary
The reader is referred to the appended independent claims. Some preferred
features are laid out in the dependent claims.
An example of the present invention is a method in a small cell base station
comprising switchable antennas of controlling reception pattern by selecting a subset
of the antennas for use. The method comprises determining a ranking value for each
of subsets of the antennas dependent upon measurements of received signal quality,
and choosing the subset of the antennas for use that provides the highest ranking
value.
Some preferred embodiments involve dynamically choosing the antenna
subset, in other words, antenna or antenna combination, that is optimum or near
optimum, in a manner that is simple, fast and responsive to changes in the radio
environment. Some preferred embodiments provide a good way to select antennas so
as to mitigate radio interference. Some preferred embodiments involve a systematic
evaluation of available antenna subsets to choose, for use, the one with a highest
ranking. Some preferred embodiments involve periodically evaluating various antenna
subsets so as to be adaptive to changes in the radio environment.
Some preferred embodiments involve intelligently selecting antenna subsets
for testing and possible use, reducing performance losses due to incorrect antenna
selection, and adaptive learning from previous evaluations of antenna subsets.
In some preferred embodiments, excessive testing of unsuitable antenna
subsets is avoided.
Some preferred embodiments function in real-time, so heavy computation is
not involved. Also, some preferred embodiments are responsive to changes in the
radio environment, such as due to user terminal mobility and the initiation of new
voice calls or data sessions.
Compared to the known approach of using an omni-directional antenna, uplink
interference to signals from the small cell-connected user terminals may be reduced.
Lower transmit powers from small cell base stations may result, giving longer battery
life and 'greener' communications. As a possible additional benefit, downlink
interference from the small cell base station to users of other base stations may also be
reduced in consequence.
Brief Description of the Drawings
Embodiments of the present invention will now be described by way of
example and with reference to the drawings, in which:
Figure 1 is a diagram illustrating a wireless communications network
according to a first embodiment of the present invention,
Figure 2 is a diagram illustrating an example femtocell base station
deployment within one macrocell shown in Figure 1,
Figure 3 is a diagram illustrating a femto with a multi-element switchable
antenna,
Figure 4 is a diagram illustrating in greater detail some circuitry in the femto
shown in Figure 3,
Figure 5 is a diagram illustrating an example of the iterative femto operating
cycle involving using a selected antenna subset (first use period), testing an antenna
subset (testing interval) to choose an antenna subset for subsequent use, then using
chosen antenna subset (second use period), and
Figure 6 is a diagram illustrating learning rate factor adaptation for an antenna
subset as a function of time since previous testing of that antenna subset.
Detailed Description
The inventors realised that implementing classic beam forming, namely the use
of multiple antenna elements with particular amplitudes and phase shifts is not
feasible in femtos.
The inventors realised that use of a multi-element switchable antenna is useful
for interference mitigation. This is where a number of antennas are provided and a
switch is used to select which antenna or combination of antennas to use. The
inventors realised that the computational complexity of this approach is generally low,
but as it depends on the number of antennas being used, use of more than two
antennas at a time is usually avoided.
The inventors realised that to make effective use of this approach, interference
must be managed by addressing how to dynamically select an antenna pattern giving a
good signal, from the user terminal of interest, with little interference.
We now describe example embodiments of the invention starting with a
description of the network before describing a femtocell base station and its
functionality. The functionality is based on cycles each involving intelligent selection
of an antenna subset to test, testing of the antenna subset, and intelligent choice of an
antenna subset for use in communications.
Network
As shown in Figures 1 and 2, a network 10 for wireless communications,
through which a user terminal 34 may roam, includes two types of base station,
namely macrocell base stations and femtocell base stations (the latter being sometimes
called "femtos"). One macrocell base station 22 is shown in Figures 1 and 2 for
simplicity. Each macrocell base station has a radio coverage area 24 that is often
referred to as a macrocell. The geographic extent of the macrocell 24 depends on the
capabilities of the macrocell base station 22 and the surrounding geography.
Within the macrocell 24, each femtocell base station 30 provides wireless
communications within a corresponding femtocell 32. A femtocell is a radio coverage
area. The radio coverage area of the femtocell 32 is much less than that of the
macrocell 24. For example, the femtocell 32 corresponds in size to a user's office or
home.
As shown in Figure 1, the network 10 is managed by a radio network
controller, RNC, 170. The radio network controller, RNC, 170 controls the operation,
for example by communicating with macrocell base stations 22 via a backhaul
communications link 160. The radio network controller 170 maintains a neighbour list
which includes information about the geographical relationship between cells
supported by base stations. In addition, the radio network controller 170 maintains
location information which provides information on the location of the user
equipment within the wireless communications system 10. The radio network
controller 170 is operable to route traffic via circuit-switched and packet-switched
networks. For circuit-switched traffic, a mobile switching centre 250 is provided with
which the radio network controller 170 may communicate. The mobile switching
centre 250 communicates with a circuit-switched network such as a public switched
telephone network (PSTN) 210. For packet-switched traffic, the network controller
170 communicates with serving general packet radio service support nodes (SGSNs)
220 and a gateway general packet radio support node (GGSN) 180. The GGSN then
communicates with a packet-switch core 190 such as, for example, the Internet.
The SC 250 , SGSN 220, GGSN 180 and operator IP network constitute a
so-called core network 253. The SGSN 220 and GGSN 180 are connected by an
operator IP network 215 to a femtocell controller/gateway 230.
The femtocell controller/gateway 230 is connected via the Internet 190 to the
femtocell base stations 30. These connections to the femtocell controller/gateway 230
are broadband Internet Protocol connections ("backhaul") connections.
In Figure 2, three femtocell base stations 30 and corresponding femtocells 32
are shown for simplicity.
It is possible for a mobile terminal 34 within the macrocell 24 to communicate
with the macrocell base station 22. When the mobile terminal 34 enters into a
femtocell 32 for which the mobile terminal is registered for communications within
the femtocell base station 30, it is desirable to handover the connection with the
mobile terminal from the macrocell to the femtocell. In the example shown in Figure
2, the user of mobile terminal 34 is a preferred user of the nearest 32' of the
femtocells 32.
As shown in Figure 2, the femtocell base stations 30 are connected via the
broadband Internet Protocol connections ("backhaul") 36 to the core network (not
shown in Figure 2) and hence the rest of the telecommunications "world" (not shown
in Figure 2). The "backhaul" connections 36 allow communications between the
femtocell base stations 30 through the core network (not shown). The macrocell base
station is also connected to the core network (not shown in Figure 2).
Femtocell base station
As shown in Figure 3, the femto includes processing circuitry 50 connected to
a backhaul interface 52 connected to a backhaul DSL line 36. The processing circuitry
is connected to transmit amplifier 58 and receive amplifier 60 that are both connected
to a diplexer 62. The diplexer 62 is connected to a switch 64. The switch 64 is a oneto-
four switch having four outputs 66 each connected to a respective antenna
(sometimes referred to as an antenna element) 68. For computational simplicity and to
keep impedance mismatches within acceptable limits, not more than two antenna
elements are connected at any one time. Accordingly for the four antenna system there
is a total of ten antenna patterns possible (in other words, ten antenna subsets, in other
words a choice of ten different antennas or antenna combinations, are possible).
Namely, these are the four antennas individually, plus six possible combinations of
two antennas (namely first and second, first and third, first and fourth, second and
third, second and fourth, and third and fourth).
In this example the antenna elements are each patch antennas being small and
ease to connect within the housing of the femto which is approximately 17 by 15 by
3.5 centimetres in size, and contains a single main circuit board(not specifically
shown). Patch antennas are flat in shape and easily located in the base station placed
parallel to the main circuit board.
As shown in Figure 4, the circuitry 50 of the femto 30, includes a processor 5 1
operative to determine received Signal-to-Interference Ratio, SIR, a comparator 53
operative to compare the determined values to a target SIR value, and a user terminal
transmit power controller 55. The circuitry 50 also includes a switch controller 57
connected to the switch 64.
Femtocell Base Station Function
At any given time the femto should be operating with a radio reception pattern
which provides high gain towards the a femto-connected user terminal and low gain
towards other user terminals, such as macrocell-connected users and users connected
to other femtos. The fitness of each pattern is judged by assessing the transmission
power required from the user terminal connected to the femto that satisfies an uplink
Signal to Interference ratio, SIR, requirement of the femto. As the femto uses
WCDMA technology, and the transmission power of the user terminal is controlled by
the femto in a fast power uplink control mechanism, the femto monitors the relative
power of its users even if the actual initial power is unknown. Accordingly the femto,
particularly its processor 51, performs frequent repeat measurement of received
Signal-to-Interference ratio, and each time, the comparator 53 compares the result to
the target SIR requirement Upon finding that the measured SIR exceeds the Target
SIR, the femto, specifically the user terminal power controller 55, commands the user
terminal to lower its transmit power. On the other hand, upon measured SIR going
below the Target SIR, the femto commands the user terminal to increase its transmit
power. Accordingly, in this way, the transmit power of the user terminal relative to the
initial transmit power of the user terminal is known to the femto.
Testing of antenna subsets
The switch controller 57 controls the switch 64 so that the femto tries possible
antenna subsets (effectively possible antenna patterns), and selects for use the most
appropriate one, namely the one requiring low transmit power to satisfy the SIR target,
else if the maximum uplink transmit power is reached but the SIR target is not
reached, the antenna subset giving the best SIR is selected for use.
The inventors realised that, on the one hand, frequent testing of antenna
subsets would result in fast adaptation in a dynamically changing radio environment
but because some of the antenna subsets might perform poorly, frequent drops in SIR
might occur. The inventors also realised that very frequent SIR measurements may not
be acceptable in view of limits for acceptable packet loss and packet delay, assuming
retransmission is allowed. The inventors realised that how to test antenna subsets was
a trade-off, aimed at providing reasonable fast adaptation but an acceptable packet
loss rate and packet delay.
Testing one antenna subset in each interval between use periods
The inventors realised that testing all antenna subsets in a testing interval
before each use period would be excessive. The inventors also realised that totally
avoiding testing antenna subsets that had performed poorly in the past would be
disadvantageous in a changing radio environment where an antenna subset might
perform better after a time. Accordingly the inventors came to the following approach.
As shown in Figure 5, a single antenna subset is tested in each testing interval
59 between use periods 61. It can be considered that there are cycles of operation,
each cycle having an antenna subset testing phase 59 followed by a use period 6 1
using the best found antenna subset so far, according to a ranking scheme explained in
more detail below.
The interval required to test a single antenna subset is one UMTS time slot
(equivalent to 0.66 milliseconds) because this is the rate at which an SIR evaluation is
possible being the time between measurement reports provided for power control in
line with the UMTS standard.
This approach can be considered as a distributed testing scheme in which
errors due to power fluctuations, and consequential SIR drops, in the uplink direction
are likely small during testing, because the testing intervals are short.
Additionally, to combat potential SIR drops down to unacceptable levels
during testing, user terminals connected to the femto are commanded to temporarily
increase their transmit power just before the testing occurs. This is described in more
detail later below.
As explained in more detail below, from the testing, the various antenna
subsets are ranked. After a smoothing operation as explained below with reference to
Figure 6, the antenna subset with the best ranking is selected and used. The rankings
are, of course, updated over time.
Ranking the antenna subsets
The ranking is dependent upon the radio environment, the user terminal's
location, and the radio channel properties. The ranking may be considered an estimate
of how well a particular antenna subset can capture the user terminal of interest and
simultaneously avoid interfering users. This means high antenna gain toward the user
terminal of interest and low antenna gain towards interfering users. Since only the the
relative values of the rankings are important, the (raw) ranking R of antenna subset x
is the difference between the observed SIR (in dB) for that antenna subset and the
uplink transmit power of the user terminal (in dBm). This is shown mathematically as
R(x)=SINRx - P p nk_
As mentioned previously, since the uplink transmission powers of user
terminals connected to the femto are controlled by the femto, the femto tracks the
transmit power of each of the user terminals. (The transmit power as tracked are
relative transmit powers, being relative to a respective initial transmit power.)
Except for the time slots at which testing occurs (the testing interval), at each
time slot the femto compares the ranking of the possible antenna subsets and selects
the one with the highest ranking for use.
Updating the ranking of antenna subsets
As mentioned previously, SIR values are calculated at each time slot for power
control of the user terminals. Hence a new estimate of ranking value is determined
every time slot for the current (in use or being tested) antenna subset.
The changes of ranking values are smoothed so that the updated (smoothed)
ranking value R(x) is the old value (R(x)oid) + weighted difference between the old
value and the estimated new (raw) value (R(x)new ). This can be considered a temporal
difference averaging method according to
R(x) <= R(x)oid + . [R(x)„ew - R(x)oid]
where a is the learning rate between 0 and 1. A higher learning rate suppresses
contributions due to older measurements faster than a lower learning rate and the
learning rate is adaptively changed dependent on the time since that antenna subset
was last tested.
As shown in Figure 6, specifically an ascending second order convex function
is used, namely
a=(N(x)/k) 2+c
where N(x) is the time in timeslots since last testing or using of antenna subset x, and
K and c are constants. Specifically, c is the minimum learning rate and K is a scaling
factor that defines the slope of the curve.
The effect is that if measurements are close together in time, averaging is
useful and a low learning rate is expected. On the other hand, if the time between
measurements is long, then the chance that older measurements are outdated
increases, so a quicker adaptation to later measurements becomes desirable and hence
a higher learning rate is used.
The inventors realised that this smoothing was appropriate because, on the one
hand, an alternative of completely replacing the old value with the new would be
undesirable because the ranking would be very susceptible to errors if based on merely
a measurement in a single time slot subject high frequency noise, but, on the other
hand, as the radio environment changes over time, fairly rapid adaptation of ranking is
desirable.
As a practical matter, if a previous measurement is older than 5 seconds (7500
UMTS time slots) the previous ranking is fully over-written completely (in other
words, a =1).
Choosing which antenna subset to test
The choice of use period duration, in other words the temporal separation
between testing intervals, is a trade-off between the rate of adaptation of the femto to
its radio environment and the risk of SIR drops due to testing of a poor performing
antenna subset. In this example, a use period of between approximately 50 and 150
UMTS time slots, in other words 0.03 to 0.1 milliseconds is appropriate. For example
one testing interval ( 1 UMTS time slot ) occurs after a use period of approximately
100 UMTS timeslots.
In each testing interval, an antenna subset other than the one currently in use is
tested. In deciding which antenna subset to test, in addition to ranking values, the
time since that antenna subset was last measured, in other words, last tested or used, is
also considered. The reason for this is that the performance of an antenna subset is
very much dependent on time-varying factors such as the user terminal location, the
local environment and radio channels used, so a poorly-performing antenna subset
may get better after a time.
In this example a simple and linear combination of the two metrics is used,
(although in other embodiments, the two metrics are combined in different ways.)
Specifically, the choice of which antenna subset to test is the one which gives the
highest sum of its Ranking and time since last tested appropriately scaled. This is
Explore (x) =R(x) +. ()
where Explore(x) is the measure of suitability of the antenna subset V for test
and N(x) is the number of time slots from the previous test, or use, of the antenna
subset '' and is a scaling factor.
The value of Explore is determined for all antenna subsets except the one that
is currently in use and the antenna subset giving the highest Explore value is selected
for testing.
Power increase before testing
There is always a degree of risk due to SIR drops during testing, even though
the intelligent choice of antenna subset to test reduces this risk. Accordingly, to
minimise the impact of SIR drops, for the testing interval the transmit power is
increased of user terminals connected to the femto. This is effected by sending a
power control command from the femto to the user terminals.
A maximum transmit power increase (P a Ex o ) is set so as to avoid
excessive interference to other user terminals, however the maximum increase is often
unnecessary. The increase in transmit power to command is calculated based on
antenna ranking, or based on statistical similarity, or based on both antenna ranking
and statistical similarity. These three options are now described in turn.
Calculating Power Increase based on Antenna Ranking
In this first approach, accurate rankings are assumed for this purpose and the
power increase to be applied is simply determined based on the difference between the
two ranking values. Specifically the power increase is the less of the maximum
allowed power increase before the test interval (P Ma E iore ) d the difference between
the ranking value of the last used antenna subset (that use being in the last use period)
and the ranking value of the antenna subset under test. This is on that assumption that
the ranking of the last used one is higher than the ranking of the one under test, which
is appropriate as the antenna subset with highest ranking is in use/ selected for use.
The above may be written mathematically as:
PExp!ore_ranking = in ([R(laSt_USed_Subset)- R(teSt_Subset)], P M ax_ExpIor)
where
R(last_used _subset)- R(test_subset) >= 0
Calculating Power Increase based on Statistical Similarity
In an otherwise similar embodiment, the statistical similarity of the gain
patterns of the two antenna subsets (the one used in the last use period and the one
under test) is used instead to determine the level of the power increase to apply. This
does not rely on an assumption that ranking values are accurate.
The idea is that if the patterns are found to be very different then, to be on the
safe side, a larger power increase is more appropriate.
In this approach, a similarity index is determined.
The similarity index i s a value in range of [0, 1] and is defined as the ratio of
the common area between the two antenna patterns divided b y the maximum of the
areas under the two patterns. In other words, if s represents the gain pattern of
the mth antenna pattern. The index of similarity between pattern n and m is defined as
In this approach the power increase before testing is set based merely on the
degree of similarity between the two patterns: the more the two antenna subsets are
dissimilar in this regard, the more should power be increased before the test. This
follows:
PExplore_statistical = [l-I(laSt_USed_Subset, teSt SUbset)] * PMax_Exp!or
where I is the similarity index.
The femto includes a look-up table in which the similarity index values are
stored. Given n individual antenna subset possibilities the femto stores n* (n-1)
similarity index values.
Calculating Power Increase based on both Antenna Ranking and Statistical Similarity
In a further otherwise similar embodiment, both the antenna ranking and the
statistical similarity of the gain patterns of the two antenna subsets are used to
determine the level of the power increase to apply.
This approach may be considered a useful compromise between the two
approaches described above, where one assumes fully accurate ranking estimations
and the other is purely based on general statistical similarity between the patterns of
the antenna subsets.
Accordingly, in this example the actual power increase is considered as a
weighted combination of those two metrics
^Explore W i * PExplore_ranking statistical * PExplore_statistical
where
an in = 1- W statistical
The weightings of each power increase component reflects the certainty level
of the ranking estimations, which is heavily dependent on the time from the previous
evaluation of the ranking of the antenna subset under test.
In its simplest form, ani ng selected to be a linearly decreasing function of
N(x), where x is the antenna subset to be tried out and with offset of 1 at start, for
example:
max( [-N(x)/T_deccor] + 1 , 0 )
where T_deccor is the decorrelation time of the system (expressed in the same
units as N(x), namely either pure time or the number time slots, in such way that if
the previous evaluations were older than T_deccor (N(x)> T decorr), the system
would merely uses statistical similarity of the patterns to determine the power increase
to be applied.
It may be considered that, in this example, the level of power increase is
determined dependent upon the difference between the last determined ranking value
of the last used subset of antennas and the last determined ranking value of the subset
being measured; and also dependent upon an index of the similarity between the
respective radio reception patterns of the two subsets of antennas. In determining the
level of power increase to apply, the difference and the index are relatively-weighted,
the weighting being dependent upon time since the antenna subset under test was last
tested.
In some example systems there are a limited number of antenna subsets (for
example ten when a femto having four antennas uses a maximum of two antennas at
any time), and the base station readily tracks ranking values and time since previous
testing of an antenna subset. Furthermore, as regards instructing an increase of the
power of user terminals for testing of an antenna subset, the femto includes a look up
table which stores the similarity index values.
General
The present invention may be embodied in other specific forms without
departing from its essential characteristics. The described embodiments are to be
considered in all respects only as illustrative and not restrictive. The scope of the
invention is, therefore, indicated by the appended claims rather than by the foregoing
description. All changes that come within the meaning and range of equivalency of
the claims are to be embraced within their scope.
In some alternative embodiments, Inverted-F-antennas (IFAs) are used in place
of patch antennas. An IFA antenna is easily placed to the upper two corners of the
circuit board or is implemented directly on the main circuit board as a printed antenna.
A person skilled in the art would readily recognize that steps of various abovedescribed
methods can be performed by programmed computers. Some embodiments
relate to program storage devices, e.g., digital data storage media, which are machine
or computer readable and encode machine-executable or computer-executable
programs of instructions, wherein said instructions perform some or all of the steps of
said above-described methods. The program storage devices may be, e.g., digital
memories, magnetic storage media such as a magnetic disks and magnetic tapes, hard
drives, or optically readable digital data storage media. Some embodiments involve
computers programmed to perform said steps of the above-described methods.
Claims:
. A method in a small cell base station comprising multiple antennas of controlling
reception pattern by selecting a subset of the antennas for use;
the method comprising determining a ranking value for each of subsets of the
antennas dependent upon measurements of received signal quality, and choosing the
subset of the antennas for use that provides the highest ranking value.
2. A method according to claim 1, in which the measurements are of the difference
between received signal to interference ratio of a signal received from a user terminal
and transmit power of the user terminal.
3. A method according to claim 2, in which the small cell base station knows the
transmit power of the user terminal as the small cell base station controls that transmit
power.
4. A method according to any preceding claim, in which, in each test interval, a
single subset of the antennas is tested and its ranking is compared to that of the one of
the subsets of antennas that was last used and then highest ranking, so as to determine
the subset of the antennas that provides the highest ranking.
5. A method according to claim 4 in which the test interval is between use periods,
the test interval is a single time slot, and each use period is at approximately 25 to 75
time slots.
6. A method according to claim 5, in which each time slot is of 0.66 milliseconds.
7. A method according to any preceding claim, in which the measurement of an
antenna subset is used to adapt a previously determined ranking value for that subset,
the weighting given to the measurement depending upon the time since last
measurement of that subset, to provide the ranking value.
8. A method according to claim 7, in which more weight is given to the
measurement where the last measurement is relative old.
9. A method according to claim 7 or claim 8, in which up to a predetermined time
since last measurement, the weight is a second order function dependent upon the time
since last measurement, the weight being between a positive minimum and a
maximum of one.
0. A method according to any preceding claim, in which the choice of which subset
of antennas to test is determined from ranking value and time since last measurement.
. A method according to claim 10, in which the subset is chosen for testing having
the highest value for the sum of its ranking value and its scaled time since last
measurement.
12. A method according to any preceding claim, in which the small cell base station
sends an instruction for user terminals connected to the small cell base station to
temporarily increase their transmit power for the measurement.
13. A method according to claim 12, in which the level of power increase is
determined dependent upon the difference between the last determined ranking value
of the last used subset of antennas and the last determined ranking value of the subset
being measured.
1 . A method according to claim 12 or claim 13, in which the level of power
increase is determined dependent upon, or also dependent upon, an index of the
similarity between the respective radio reception patterns of the two subsets of
antennas.
A small cell base station comprising multiple antennas,
a controller configured to control reception pattern by selecting a subset of the
antennas for use,
a processor configured to determine a ranking value for each of subsets of the
antennas dependent upon measurements of received signal quality, and
a selector configured to choose the subset of the antennas for use that provides
the highest ranking value.