METHOD AND DEVICE FOR SELECTING USER TERMINAL
SO AS TO ENHANCE RECIPROCITY ERROR CALIBRATION
BETWEEN UPLINK AND DOWNLINK
FIELD OF THE INVENTION
The present invention relates to communication field, and more particularly
relates to base station and mobile station in time division duplex system.
BACKGROUND OF THE INVENTION
Channel reciprocity character, namely the symmetry character between
uplink (UL) and downlink (DL) frequencies, has good application prospect in LTE-A
(Advanced - Long Term Evolution) TDD (Time Division Duplex) system. The
industry commonly accepts the assumption of reciprocity between uplink and
downlink, and uses the assumption to estimate the channel effectively. However, in
practice, for the reason of the difference of RF (Radio Frequency) circuits between the
receiver and the transmitter, especially the difference of RF circuits between the
receiver and the transmitter at base station side, the reciprocity between uplink and
downlink in TDD system is very hard to be guaranteed. For the TDD based system
with reciprocity between uplink and downlink, the system performance is highly
sensitive to uplink/downlink channel reciprocity errors, slight reciprocity errors
between uplink and downlink might cause significant performance degradation.
Therefore, reciprocity calibration in TDD system attracts the interest of the industry.
The calibration for OTA (Over The Air) interface of TDD system with
reciprocity between uplink and downlink becomes an effective way to guarantee
reciprocity between uplink and downlink of TDD system, since there is no need to
introduce additional hardware devices. However, the accuracy of OTA calibration
mainly depends on channel estimations between uplink and downlink, and thus how
to perform user terminal (also called user equipment or UE) selection is an important
problem, especially for CoMP (Coordinated Multi Point) system, which user
terminals are selected for calibration is a problem urgently to be solved.
Exiting time division duplex technique proposes that the selected user
terminals should locate near the center of coordinated multi-point cell cluster, which
means only user terminal locations are taken into consideration for user terminal
selections.
SUMMARY OF THE INVENTION
Because the selections of user terminal according to user terminal locations
only consider the large scale fading. However, considering shadow effect and fast
fading, the above-mentioned selection way is inaccurate.
Therefore, the present invention provides a method and corresponding device
for selecting user terminal for calibration for reciprocity error between uplink and
downlink, so as to enhance the accuracy of calibration for reciprocity error between
uplink and downlink in TDD system. The solution of the present invention is suitable
for single cell scenario and multi-cell CoMP scenario.
According to the first aspect of the present invention, there is provided a
method of selecting user terminal so as to enhance calibration for reciprocity error
between uplink and downlink, in a network equipment in time division duplex
communication systems, comprising the following steps: obtaining channel quality
related information between a plurality of first user terminals and a first base station,
wherein said channel quality related information is used for indicating channel
qualities between said plurality of first user terminals and said first base station;
selecting, at least one first user terminal as a calibration user terminal for calibrating
said reciprocity error between uplink and downlink, according to said channel quality
related information.
According to the second aspect of the present invention, there is provided a
device for selecting user terminal so as to enhance calibration for reciprocity error
between uplink and downlink in a network equipment in time division duplex
communication systems, comprising: an obtaining means, for obtaining channel
quality related information between a plurality of first user terminals and a first base
station, wherein, said channel quality related information is used for indicating
channel qualities between the plurality of first user terminals and said first base station;
a selecting means, for selecting, at least one said first user terminal as calibration user
terminal for calibrating said reciprocity error between uplink and downlink, according
to said channel quality related information.
The best user terminal may be selected for calibration for reciprocity error
between uplink and downlink, with the solution of the present invention. For CoMP
scenario, different calibration user terminals may be utilized to realize that any two
base stations (BS) in coordinated cell cluster are connected so that the entire cell
cluster calibration weights are obtained.
BRIEF DESCRIPTION OF THE DRAWINGS
By reading the detailed description of the non-limiting embodiments with
reference to the following drawings, other features, objects and advantages of the
present invention will become apparent.
Fig. 1(a) and Fig. 1(b) respectively show the schematic diagrams of
topological structure of network according to an embodiment of the present invention;
Fig.2 shows a flow diagram of system method according to an embodiment
of the present invention;
Fig. 3(a) and Fig. 3(b) show different connection diagrams of coordinated
cluster composed of five base stations according to embodiments of the present
invention;
Fig.4 shows the block diagram of device according to an embodiment of the
present invention.
Wherein, same or similar reference signs refer to the same or similar step
features or device/module.
DETAILED DESCRIPTION OF EMBODIMENTS
Firstly, radio frequency mismatch model is briefly described.
There is great difference in RF (Radio Frequency) circuits between a wireless
transmitter and wireless receiver. By precluding the effect of antenna coupling, the
effective channel responses of RF antenna lineups ( br , bt ,H , H m ) can be
modelled as diagonal matrices. Wherein, br , bt , H m , respectively denote
response of base station at receiver (RX), response of base station at transmitter (TX),
response of user terminal at receiver (RX), response of user terminal at
transmitter(TX). For example, response of base station receiver may be represented as
0 0
0 0 , where N is the number of base station antennas. The above
0 0 hbr,N
diagonal matrices are expressed as h r (t, P,T) □ Ahr (t, e t PT) , wherein A
denotes amplitude, and denotes phase.
In the reciprocity simulation, one of the most concerned parameters is the
ratio of the uplink channel response to the downlink channel response (or vice versa).
In general, if the delay of uplink/downlink transmission is neglected, the channel of
OTA (Over The Air) interface may be considered as reciprocity. However, the
mismatch of antenna lineups will cause that the effective channels (the effective
channel from the baseband of transmitter to the baseband of receiver) of unlink and
downlink do not satisfy the reciprocity. Based on the previous RF lineup response
expression, the reciprocity error between the TX and RX antenna lineups of the base
b t P,T) Ab t P,T)e t P ) . station can be expressed as wherein denotes
a diagonal entry on the reciprocity error matrix .
Therefore, The response of TX antenna lineup may be calculated from the
RX antenna lineup and the reciprocity error matrix,
br
Because both E and br are diagonal matrices, E andH .are commutable. In
most reciprocity simulations, the absolute RF antenna lineup responses might be not a
real performance factor. To the contrary, the relative difference values between TX
and RX antenna lineups, such as the reciprocity error matrix E , might play the real
role.
All above description is about the base station side. Extension of the model to
the mobile station side is straightforward. By including the non-reciprocity effect of
both the base station side and the user terminal side, we have,
Wherein H = T
UL . By applying the reciprocity error definitions
to the above equations, the effective downlink and uplink channel may
= i ,„,
be related as,
This equation can be used to model the reciprocity of the effective uplink and
downlink RF channels. OTA interface calibration is a solution of ensuring channel
reciprocity, by which the error E 1 and E can be obtained from the uplink and
downlink CSI (Channel State Information). The uplink CSI is obtained from uplink
estimation, while the downlink CSI is fed back by the calibration user terminals.
Usually for downlink transmission, E 1 influences less, and thus it is also possible to
only make calibration for E of TX.
Fig. l a shows a schematic diagram of topological structure of network
according to an embodiment of the present invention, in which base station l a locates
in a single cell system, and Fig. lb shows another schematic diagram of topological
structure of network according to an embodiment of the present invention, in which
base station l a locates in multi-cell CoMP system. In Fig. lb, base stations la, lb and
l c coordinately serve a plurality of user terminals 2a, 2b, 2c, 2d, 2e and 2f. In Fig. lb
only three coordinately working base stations la, lb and l c and six user terminals 2a,
2b, 2c, 2d, 2e and 2f are shown. Those skilled in the art may understand that the
coordinately working base stations are not limited to the above-mentioned base
stations, or it is also possible that there are only two coordinately working base
stations, and the user terminals are not limited to the six terminal users shown in
Fig. lb. Further, those skilled in the art should also understand that whether the base
station l a works at single cell mode or at CoMP mode has been determined during
initial network planning and deployment. In the following network operation
procedure, the base station l a works according to the configuration during network
planning phase.
Fig.2 shows a flow diagram of system method according to an embodiment
of the present invention. Firstly, in step S20, each of base stations la, lb and l c
respectively obtains channel quality related information between a plurality of user
terminals and the base station, in which the channel quality related information
between base station and user terminals may be the channel information of uplink
estimation as well as downlink channel information fed back by the user terminals.
The channel quality information including uplink channel SINR(Signal Interference
Noise Ratio) information estimated by base stations la, lb or l c according to uplink
sounding signal transmitted by each of user terminals will be taken as a example to
describe.
Then in step S21, base station l a firstly judges whether the base station
works at single cell mode or at CoMP mode. Via configuration of the network
parameters, base station l a knows that it works at single cell mode, and then the
method will go into step S22', base station l a selects at least one candidate user
terminal from a plurality of first user terminals, according to the channel quality
related information from a plurality of the first user terminals served by the base
station la, for calibrating said reciprocity error between uplink and downlink, wherein
the channel quality between the selected at least one candidate user terminal and base
station l a is greater than a first predetermined threshold .
The first predetermined threshold may be the parameters configured by the
system at the beginning of network configuration, for example, the first predetermined
threshold is configured as 16dB. SINRs between base station l a and user terminals
2a, 2b, 2c and 2d are respectively 18dB, 22dB, 25dB and 18dB. Therefore, the
channel qualities between the base station l a and user terminals 2a, 2b, 2c and 2d are
greater than the first predetermined threshold , and the base station l a takes user
terminals 2a, 2b, 2c and 2d as candidate user terminals. Here for the reason of
briefness, Table. 1 only shows the SINR values between each of user terminals and the
corresponding base stations greater than the threshold.
Then, in step S23', the base station l a compares the number of the candidate
user terminals with the predetermined number MaxUE, and the lesser value is taken
as the final number of calibration user terminals. Namely, the base station l a judges
taking X user terminals as the final calibration user terminals, in which X=
min(MaxUE, QualifiedUE), and QualifiedUE denotes the number of selected
candidate user terminals, namely, the channel qualities between these candidate
user terminals and the base station l a are greater than the first predetermined
threshold . That a plurality of user terminals having good channel qualities with the
base station l a feed channel information back to the base station l a may enhance the
calibration robustness. However, if too many user terminals feed back channel
information so as to be used for calibration, it will cause relatively huge uplink
feedback overhead and thus reduce the spectral efficiency. Therefore, limiting the
predetermined number MaxUE is beneficial to save the uplink feedback overhead.
Then, in step S24', the base station l a judges whether X is greater than 0 or
not. if X is greater than 0, the index flag will be determined as 1, which means that at
least one user terminal for calibrating the base station is found, and the selected user
terminal indices are outputted; otherwise the index flag will be determined as 0, which
means no user terminal for calibrating the base station is found.
If in step S21 base station l a knows that it works at CoMP mode via the
configuration of the network parameters, namely, Z base stations including base
station l a work coordinately, in the following Z = 3 is taken as a example to describe,
namely, there are three base stations la, lb and l c in coordinated cell cluster. Then
the method will go into step S22, each of the base stations la, lb and l c selects at
least one candidate user terminal from these user terminals according to the channel
quality related information between itself and each of user terminals respectively
obtained by each of base station, wherein, the channel qualities between these selected
candidate user terminals and base station are greater than the second predetermined
threshold.
For example, the second predetermined threshold K = 16dB. Similarly to the
example at single cell mode, SINRs between base station l a and user terminals 2a, 2b,
2c and 2d are respectively 18dB, 22dB, 25dB and 18dB; further, SINRs between base
station lb and user terminals 2b, 2e and 2f are respectively 21dB, 18dB and 17dB;
SINRs between base station l c and user terminals 2d, 2e and 2f are respectively 18dB,
21dB and 17dB; the details are shown in the following table:
Table 1
Wherein, the channel qualities between the base station l a and user terminals
2a, 2b, 2c and 2d are greater than the second predetermined threshold K; the channel
qualities between the base station lb and user terminals 2b, 2e and 2f are greater than
the second predetermined threshold K; the channel qualities between the base station
lc and user terminals 2d, 2e and 2f are greater than the second predetermined
threshold K. Each of base stations la, lb, l c respectively generates candidate set . ,
i=\,...,Z , and each candidate set comprises the user terminal, wherein, the channel
qualities between the user terminals and the base stations are greater than the second
predetermined threshold. For example, candidate set generated by the base station
l a is Ai={UE 2a, UE 2b, UE 2c, UE 2d}, candidate set A2 generated by the base
station lb is A2={UE 2b, UE 2e, UE 2f}, candidate set A3 generated by the base
station l c is A3={UE 2d, UE 2e, UE 2f}. Then, the base station l b reports the user
terminal indices of user terminals 2b, 2e and 2f and the base station index to the
base station la, wherein, the channel qualities between the base station and the
user terminals 2b, 2e and 2f are greater than the second predetermined threshold
K, and the base station l c reports the user terminal indices of user terminals 2d,
2e and 2f and the base station index to the base station la, wherein, the channel
qualities between the base station and the user terminals 2d, 2e and 2f are greater
than the second predetermined threshold K.
Then, in step S23, the base station l a finds out the common qualified
elements for each base station pair from each of sets A1 A2 and A3 and put these
common qualified elements into CommonUEset k , namely, finds out
CommonUEsetk =Ai A . k =1,2,...,CZ
2 , in which Cz
2 =Z(Z - l)/ 2 1S a
combination number, which represents all possibilities of selecting two different
elements from the set comprising Z elements. For example,
CommonUEseti=A n A2={UE 2b}, CommonUEset2= n A3={UE 2d} and
Common UEset3=A2 n A3= {UE 2e, UE 2f}.
Then, for each CommonUEset k, the base station l a selects, a user terminal
with the largest SINR average value for the base station pair corresponding to the set,
in the set, and the selected user terminal is put into a vector d , and the base station
pair corresponding to the selected user terminal is recorded. For example, for
CommonUEset 3 , the average SINR of user terminal 2e for base stations lb and l c is
(18+21)/2=19.5 dB , and the average SINR of user terminal 2f for base stations lb
and l c is (17+17)/2=17 dB , therefore, user terminal 2e is the user terminal having the
largest SINR average value for the base station pair corresponding to the set in the set,
and thus the base station l a selects the user terminal 2e and records the corresponding
base station pair, that is to say, the base stations lb and lc. Further, similarly, the
selected elements in the vector d further comprise the user terminal 2b
corresponding to the base stations l a and l b and the user terminal 2d
corresponding to base stations l a and lc.
Then, the base station l a sorts vector d by ascending according to the
average SINR between each of base station pairs and user terminals. Because the
average SINR of user terminal 2b for base station pair l a and lb is 21.5 dB , and the
average SINR of user terminal 2d for base station pair l a and l c is 18 dB , and the
average SINR of user terminal 2e for base station pair lb and l c is 19.5 dB , the result
after sorting is {UE 2d , UE 2e , UE 2b }.
Then, in step S24, the base station l a firstly initializes candidate user
terminal set S by S = d , and then initializes n=l: length (d), which means firstly n=l,
and then n adds 1 one by one till n= length (d) (the length of vector d).
Then, in step S25, the base station l a judges whether n is lesser than length
(d )+l or not, if n is lesser than length (d )+l, the method goes into step S26, the base
station l a removes the element from the set S, and then in step S27 it judges the
connectivity of the adjacent matrix B after removing the n element. In order to
describe the principle of connectivity in the following, without loss of generality, a
CoMP coordinated cell cluster comprising five base stations is taken as example to
describe.
Firstly, four candidate user terminals are selected, in which each of selected
candidate user terminals respectively relates to two base stations of the CoMP cell
cluster, namely, the channel qualities between each of the candidate user terminals
and the two base stations are both greater than the second predetermined threshold K.
Therefore, five base stations of the CoMP cell cluster are divided into four subsets. If
the subsets are connected, for example, {1,2}, {2,4}, {3, 5}, {4,5}, as shown in
Fig. 3(a), namely, any one base station may be directly or indirectly connected to
any other base stations, then, partial calibration weights are connected, and thus
the overall calibration weights of the entire CoMP cluster may be obtained.
However, as shown in Fig. 3(b), if the subsets of the base stations are
unconnected, for example, {1,2 },{1,3 },{2,3 },{4,5 }, to be specific, the base stations
4 and 5 are isolated from the base stations 1, 2 and 3, the overall calibration weights
of the entire CoMP cluster are not able to be obtained.
The corresponding adjacent matrices Bc and BunC of the two CoMP clusters are
respectively listed below.
The adjacent matrices Bc corresponding to the connected graph shown in
0 1 0 0 0
1 0 0 1 0
Fig.3 (a) is , 0 0 0 0 1 (1), and the adjacent matrices BunC corresponding to
0 1 0 0 1
0 0 1 1 0
the connected graph shown in Fig.3 (b) is i 1 1 0 0 0 (2).
0 0 0 0 1
0 0 0 1 0
The set of nodes and branches in network is called as graph, denoted as G=
(V, E), in which V denotes the set of nodes, V= {v l v2,...,v m}, m is the number of
nodes, which is usually written as m= |v| ; E denotes the set of branches, E=
{e^e^...^}, n is the number of branches, n=|E| . The two nodes corresponding to the
branch are u and v, when the flow direction of fluid is irrelevant to the studied
question, graph G is called as undirected graph.
If there is at least one path between any two nodes in graph G, the two nodes
are called as connected and the graph G is called as connected graph. For undirected
graph G= (V, ), m= square matrix B= (b ) the matrix B is
called as node adjacent matrix of graph G.
B to the power of k is denoted as Bk = (b )m m , in which (bf ) =
ft=l
For matrix S = s v ) = Bk (3), if all elements in matrix S are nonzero elements,
the graph G is connected graph; otherwise if there are T zero elements in matrix S, the
graph G is unconnected graph, in which B is the node adjacent matrix of graph G.
For the equation (1) and (2), corresponding to c in equation (1),
Therefore, the
graph corresponding to graph 3(a) is connected graph, and the graph corresponding to
graph 3(b) is unconnected graph.
Back to the above-mentioned CoMP cluster comprising three base stations la,
lb and l c and six user terminals 2a-2f. Wherein, in step 23, the result after sorting is
that vector d is {UE 2d, UE 2e, UE 2b}. And then, in step 26 the base station l a
firstly removes the first element UE 2d in d, namely, the user terminal with the worst
average channel qualities with two base stations desired to calibrate. Then, in step 27,
the base station l a judges whether the remaining elements in d make the graph
comprised by the base stations in CoMP cluster connected. Because the remaining
two elements guarantee the cluster comprised by base stations la, lb and l c
connected, then the method goes into step S29, 1 is added to n, then, the method
returns to step S25 and whether n is lesser than length(d) + 1 will be judged, if yes,
step S26 is repeated, the base station l a further removes the first element according to
the sorting in the rest of the vector d , namely the user terminal having the worst
average channel qualities with the selected base stations in the rest of vector d ,
namely UE 2e. Then, the base station l a further judges whether the remaining
elements in d make the graph comprised by the base stations in cell cluster connected.
Because only one element remains, UE2b, which is not able to make the base stations
la, lb and l c connected, and thus, in step 28, the base station l a puts UE 2e back into
the vector d . Then, the method goes into step S25, and the judging result is no, and
thus the method goes into step S26'.
In the judging procedure of step S26', if the adjacent matrix B is judged as
connected, then in step S27' the base station l a sets the index flag as 1, and the index
1 indicates that enough qualified user terminals are found, and the set S at this time
and corresponding base station pair index are outputted. For example, in
above-mentioned step S26', the base station l a finds UE 2e and UE 2b can make the
base stations la, lb and l c connected, and thus the base station l a takes the set S{UE
2e, UE 2b} as candidate user terminals and outputs the corresponding base station
pairs {the base station lb, the base station lc} and {the base station la, the base
station lb}; otherwise, the method goes into step S27', the base station l a sets the
index flag as 0, and the index 0 indicates that no enough qualified user terminals is
found.
In above-mentioned embodiment, the channel quality related information
comprises SINR information of the uplink channel obtained by the base stations la,
lb and lc. Because the estimated uplink channel information might not perfectly
match with actually required downlink channel information, alternatively, the step
S29' is further included after the step S26' and before the step S27', that is, the base
station l a indicates the above selected user terminals 2e and 2b to feed back downlink
CSI(Channel State Information). Alternatively, in step S29", the base station l a
re-judges the downlink channel information fed back by the user terminals 2e and 2b
according to the downlink channel information fed back by the calibration user
terminals 2b and 2d, for example, whether SINR is greater than the above-mentioned
second predetermined threshold. When the downlink channel information fed back by
the calibration user terminals 2e and 2b are both greater than the second
predetermined threshold, the method goes into step S27', the base station l a will
obtain calibration weights of the entire cluster according to the set S {UE 2e, UE
2b}comprising the selected calibration user terminals, the downlink channel state
information fed back by the calibration user terminals 2e and 2b, the estimated uplink
channel state information between the user terminal 2e and 2b, the corresponding base
station pairs {the base station lb, the base station lc} and {the base station la, the
base station lb}, so as to make the calibration between uplink and downlink in TDD.
Those skilled in the art may understand that the above-mentioned method is
also suitable for the single cell scenario, namely, in step S22', if the base station l a
works at single cell mode, the channel quality related information with user terminals
2a, 2b, 2c and 2d obtained by the base station l a are all uplink channel estimation
information, then in a variable embodiment, after the step S23' that the base station l a
selects the calibration user terminals, it may also indicate the selected user terminals
to measure and feed back downlink channel information. Then, the base station l a
makes the calibration between uplink and downlink according to the downlink
channel information fed back by the candidate user terminals and the uplink channel
information estimated by the base station.
In above-mentioned embodiments, SINR is taken as an example to describe
the channel quality information. In optional embodiments, the channel quality
information may also include other channel quality measurement parameters
influencing the channel estimation accuracy.
In above-mentioned each of embodiments, the implementing subject of each
step is the base station la, namely, the base station l a acts as the first base station, and
the base station lb and l c act as the second base station neighbouring the first base
station, and they report the respectively collected channel quality related information
to the base station la. Those skilled in the art can completely understand that the
present invention is not limited to the above-mentioned network structure. For
example, the base stations la, lb and l c may report their respective channel quality
related information to a management network apparatus, which comprises but is not
limited to RNC (Radio Network Controller), which manages the base stations la, lb
and lc, namely, the management network apparatus replaces the base station l a as the
execution subject of the corresponding steps shown in Fig.2.
The selections of the first predetermined threshold and the second
predetermined threshold are the same in the above context. Certainly, that the
above-mentioned first predetermined threshold and the second predetermined
threshold have the same value is only for the purpose of convenience for description.
Those skilled in the art may understand, in practice, the above-mentioned two
thresholds may have the same or different values, as well as values different from
above-mentioned example according to particular system configurations and the
settings from network administrators.
In above-mentioned embodiments, that each user terminal calibrates two base
stations is taken as example to describe. Certainly, without loss of generality, the
number of base stations which each user terminal may calibrate is not limited to two,
for example three. Then, in step S23, the expression of CommonUEset k is
accordingly modified to CommonUEset = Aj n Aq , k =l,2,...,Cl , in which
3 = Z(Z -l)(Z -2)
z 2x3
In above-mentioned embodiments, one user terminal is once selected to
calibrate one base station pair so that the feedback overhead is the least. In another
variable embodiment, if a plurality of user terminals are once selected to calibrate
each base station, in step S23, for each base station pair, namely CommonUEset k, a
plurality of user terminals having the best average channel qualities with the base
station pairs are selected and the selected plurality of user terminals are putted into the
matrix D. For example, still referring to Table 1, the best two user terminals {UE 2e,
UE 2f} for the base station pair {the base station lb, the base station lc} are taken out
and putted into the matrix D. In the following the elements in matrix D are sorted
according to the average SINR of user terminal groups. Because average SINR of the
UE 2e and UE 2f for the base station pair lb and l c is (18+21+17+17)/4 = 18.25dB,
the average SINR of the user terminal subset {UE 2e, UE 2f} is greater than the
average SINR 18dB of the user terminal 2d corresponding to the base station pair l a
and lc, and is lesser than the average SINR 21.5dB of the user terminal 2b
corresponding to the base station pair l a and lb. Therefore, in step S24, let S=D, and
thus in step 26, UE 2d is removed from the set S first, and then the judgement in step
S27 is performed, 1 is added to n in step S29, the judgement in step S25 is performed,
then in step S26, the user terminal subset {UE 2e, UE 2f} is removed from the set S,
and the subsequent judgement is performed. Because the subsequent steps are similar
to above described embodiments, which is not repeated. In the embodiment, because
each base station pair uses a plurality of user terminals to calibrate, the robustness is
further improved.
Certainly, the values listed in each of above-mentioned embodiments are
only exemplary. Those skilled in the art may understand that the above-mentioned
parameters may be measured completely in real-time in actual system operation, and
the above-mentioned parameters might also be different for different topological
structure of network.
The three sites shown in Fig.l (b) locate in a same hexagonal cellular cell,
but those skilled in the art may fully understand that the base station desired to be
calibrated may respectively locate in different cellular cells.
Hereinbefore, the present invention is described from the aspect of method
flow; the present invention will be described from the aspect of device hereinafter.
A device 40 shown in Fig.4 locates in the base station l a shown in Fig.l (a)
and Fig.l (b), certainly the device 40 may also locate in the base stations lb, lc.
The device 40 comprises an obtaining means 400, and a selecting means
401.The selecting means 401 further comprises a removing means 4010 and a first
judging means 4011.
Firstly, the obtaining means 400 obtains channel quality related information
between a plurality of first user terminals and a first base station la, wherein, the
channel quality related information is used for indicating channel qualities between
the plurality of first user terminals and the first base station.
Then, the selecting means 401 selects at least one first user terminal as
calibration user terminal for calibrating reciprocity error between uplink and downlink,
according to the channel quality-related information.
In particular, when the base station l a works at single cell mode, the
selecting means 401 firstly selects at least one candidate user terminal from the
plurality of first user terminals according to the channel quality related information,
wherein the channel quality between the at least one candidate user terminal and the
first base station l a is greater than a first predetermined threshold. The Table 1 with
the first predetermined threshold =16dB is still taken as example to describe. For
example, the user terminals 2a, 2b, 2c and 2d are selected. And then, the selecting
means 401 judges whether the number of the at least one candidate user terminals 2a,
2b, 2c and 2d is greater than the predetermined number MaxUE.
The MaxUE user terminal which is the top MaxUE user terminal as to the
channel quality with the first base station in the at least one candidate user terminal is
taken as the calibration user terminal, when the number of the at least one candidate
user terminal is greater than MaxUE. For example, the number MaxUE is set to
2 ,then the top 2 user terminals UE 2c and UE 2b in the at least one candidate user
terminal are taken as calibration user terminal.
The at least one candidate user terminal is taken as the calibration user
terminal when the number of the at least one candidate user terminal is less than or
equal to the predetermined number MaxUE. For example, the number MaxUE is set
to 5 , candidate user terminals UE 2c, UE 2b, UE 2a and UE 2d are all taken as
calibration user terminals.
In another embodiment, when the first base station l a and at least one
neighboring second base station constitute a coordinated base station cluster and work
coordinately, for example as shown in Fig. 1(b), the neighboring second base stations
include the base stations lb and lc, the obtaining means 400 is further used for
obtaining at least one first user terminal and a plurality of second user terminals,
wherein the channel quality between the at least one first user terminal and the first
base station l a is greater than a second predetermined threshold and the channel
qualities between the plurality of second user terminals and the at least one second
base station lb or l c are greater than the second predetermined threshold.
Then, the selecting means is further used for taking the user terminal among
the plurality of first user terminal and the plurality of second user terminals as
candidate user terminal, wherein the channel qualities between the candidate user
terminal and at least two base stations are greater than the second predetermined
threshold; then, taking the minimum number of the candidate user terminals which
make the coordinated base station cluster connected as the calibration user terminal,
wherein the coordinated base station cluster is connected when there is no nonzero
element in S = ¾ = B , wherein B denotes node adjacent matrix, B denotes B
to the power of k, Z denotes total number of base stations in the coordinated base
station cluster.
To be detailed, the selecting means 401 further comprises:
a removing means 4010, for removing at least one candidate user terminal
which has minimal average channel quality one after another according to average
channel quality of the candidate user terminals, wherein the average channel quality
of candidate user terminals comprises the average value of channel qualities between
the candidate user terminal and at least two base stations, and the channel qualities
between the candidate user terminal and the at least two base stations are both greater
than the second predetermined threshold;
a first judging means 4011 used for judging whether the rest of the plurality
of candidate user terminals can make the coordinated base station cluster connected;
the removing means 4010 and the first judging means 4011 repeat the
above-mentioned operation till the rest of the plurality of candidate user terminals
can't make the coordinated base station cluster connected, when the rest of the
plurality of candidate user terminals can make the coordinated base station cluster
connected;
then, the selecting means 401 is further used for taking the minimal number
of the rest of candidate user terminals which can make the coordinated base station
cluster connected as the calibration user terminal.
In a variable embodiment, when the channel quality comprises SINR
information , the obtaining means 400 is further used for:
obtaining downlink CQI (Channel Quality Indication) fed back by the
plurality of first user terminals, and then obtaining the SINR information according to
the downlink CQI.
In a variable embodiment, when the base station l a is in single cell scenario,
the channel quality further comprises uplink channel estimation information obtained
by the base station, then, the obtaining means 400 is further used for obtaining
downlink channel quality information fed back by the calibration user terminal; then,
the device 40 further comprises a second judging means (not shown in Fig.4), for
judging whether the downlink channel quality fed back by the calibration user
terminal is greater than the first predetermined threshold; the device 40 is further used
for calibrating the reciprocity error between uplink and downlink according to the
calibration user terminal, when the downlink channel quality fed back by the
calibration user terminal is greater than the first predetermined threshold; otherwise,
the device 40 can not calibrate the reciprocity error between uplink and downlink
according to the calibration user terminal.
In a variable embodiment, when the base station l a is in CoMP scenario, and
the channel quality comprises uplink channel estimation information obtained by the
base station, the obtaining means 400 is further used for obtaining downlink channel
quality information fed back by the calibration user terminal. Then the device 40
further comprises: a third judging means (not shown in Fig.4) forjudging whether the
downlink channel quality fed back by the calibration user terminal is greater than the
second predetermined threshold; the device 40 is further used for calibrating
reciprocity error between uplink and downlink according to the calibration user
terminal, when the downlink channel quality fed back by the calibration user terminal
is greater than the second predetermined threshold.
The simulation result according to the present invention is presented in the
following. Firstly, the simulation parameters are shown in the following Table.2:
Table 2
The simulation results of single cell MU-MIMO (Multi-User Multi-Input
Multi-Output) are shown in Table 3, where the top 3 user terminals are selected as the
calibration user terminals. The average throughput only drops 1.48% by using the
OTA calibration present by the present invention. The hardware self calibration
results are also shown in Table 3: a (IdB, lOdeg) residual error causes 24.8%
throughput loss, while a (0.5dB, 5deg) residual error causes 7.73% throughput loss.
The hardware self calibration accuracy is highly depending on the RF circuitry, which
means that more cost is necessary for high accuracy.
Table 3
In Table 4, the simulation results for intra-site CoMP are shown. The same
residual error (IdB, lOdeg) reduces the throughput about 41.97%. It is obvious that
CoMP is more sensitive to reciprocity error compared with single cell MU-MIMO.
The user terminal selection scheme of the present invention is very efficient
for this scenario. If the single cell best SINR scheme is simply introduced for CoMP
calibration user terminal selection, the throughput will drop about 36.75%. However,
the average throughput only drops 23.70% by using the connected partial Best SINR
scheme for CoMP in the preferable embodiment of the present invention. In addition,
two user terminals may be used for calibration, and the average throughput is
improved 13%.
Table 4
It should be noted that the above-mentioned embodiment is only illustrative
rather than limitation to the present invention. Any technical solutions without
departing from the spirit of the present invention should fall within the protection
scope of the present invention, which comprises: the different technical features used
in different embodiments may be combined with each other so as to obtain beneficial
effect. In addition, any reference signs in the claims should not be regarded as limiting
the related claims; the term "comprise" does not exclude the device or steps not listed
in other claims or descriptions; the term "a/an" preceding an element does not exclude
a plurality of such element exist; in a device comprising a plurality of means, the
function of at least one of the plurality of means may be implemented by the same
hardware or software module; the terms such as "first", "second", "third" are used to
represent name rather than any specific order.
CLAIMS
1. A method of selecting user terminal so as to enhance calibration for
reciprocity error between uplink and downlink, in a network equipment in time
division duplex communication systems, comprising the following steps:
A. obtaining channel quality related information between a plurality of first user
terminals and a first base station, wherein said channel quality related information is
used for indicating channel qualities between said plurality of first user terminals and
said first base station;
B. selecting, at least one first user terminal as a calibration user terminal for
calibrating said reciprocity error between uplink and downlink, according to said
channel quality related information.
2. A method according to claim 1, wherein when said first base station works at
single cell mode, said step B further comprises:
Bl. selecting at least one candidate user terminal from said plurality of first user
terminals according to said channel quality related information, wherein the channel
quality between said at least one candidate user terminal and said first base station is
greater than a first predetermined threshold;
B2. judging whether the number of said at least one candidate user terminal is
greater than a predetermined number N:
- taking N said user terminal which is the top N user terminal as to the channel
quality with said first base station in said at least one candidate user terminal as said
calibration user terminal, when the number of said at least one candidate user terminal
is greater than N;
- taking said at least one candidate user terminal as said calibration user terminal,
when the number of said at least one candidate user terminal is less than or equal to N.
3. A method according to claim 1, wherein when said first base station and at
least one neighboring second base station constitute a coordinated base station cluster
and work coordinately, said step A further comprises:
- obtaining at least one first user terminal and, a plurality of second user
terminals, wherein, the channel quality between said at least one first user terminal
and said first base station is greater than a second predetermined threshold, and the
channel qualities between said plurality of second user terminals and said at least one
second base station are greater than said second predetermined threshold;
said step B further comprises:
.taking at least one user terminal among said at least one first user terminal
and said plurality of second user terminal as candidate user terminal, wherein the
channel qualities between said at least one user terminal and at least two base stations
are greater than said second predetermined threshold;
B2'. taking a plurality of said candidate user terminals which make said
coordinated base station cluster connected as said calibration user terminal, wherein,
said coordinated base station cluster is connected when there is no nonzero element in
5 = (5,j ) = B ,wherein, B denotes node adjacent matrix, B denotes the Kth
power of B, Z denotes total numbers of base stations in said coordinated base station
cluster.
4. A method according to claim 3, wherein said step B2' further comprises:
2Ί . removing at least one candidate user terminal which has minimal average
channel quality one after another according to average channel quality of said
candidate user terminals, wherein said average channel quality of candidate user
terminals comprises the average value of channel qualities between said candidate
user terminal and at least two base stations, and the channel qualities between said
candidate user terminal and said at least two base stations are both greater than said
second predetermined threshold;
B2'2. judging whether the rest of the plurality of candidate user terminals can
make said coordinated base station cluster connected;
B2'3. repeating said step B2' land B2'2 till the rest of a plurality of candidate
user terminals can't make said coordinated base station cluster connected, when said
rest of said plurality of candidate user terminals can make said coordinated base
station cluster connected;
- taking the minimal number of said rest of candidate user terminals which can
make said coordinated base station cluster connected as said calibration user terminal.
5. A method according to any one of claims 1 to 4, wherein when said channel
quality comprises signal to interference plus noise ratio information, said step A
further comprises:
Al. obtaining downlink channel quality indication fed back by said plurality of
first user terminals;
A2. obtaining said signal to interference plus noise ratio information according to
said downlink channel quality indication.
6. A method according to claim 2, wherein when said channel quality comprises
uplink channel estimation information obtained by said base station, after said step B,
the method further comprises:
- obtaining downlink channel quality information fed back by said calibration
user terminal;
- judging whether said downlink channel quality fed back by said calibration user
terminal is greater than said first predetermined threshold;
- calibrating reciprocity error between uplink and downlink according to said
calibration user terminal when said downlink channel quality fed back by said
calibration user terminal is greater than said first predetermined threshold,.
7. A method according to claim 3, wherein when said channel quality comprises
uplink channel estimation information obtained by said base station, wherein after
said step B, the method further comprises:
- obtaining downlink channel quality information fed back by said calibration
user terminal;
- judging whether said downlink channel quality fed back by said calibration
user terminal is greater than said second predetermined threshold;
- calibrating reciprocity error between uplink and downlink according to said
calibration user terminal, when said downlink channel quality fed back by said
calibration user terminal is greater than said second predetermined threshold.
8. A method according to any one of claims 1 to 7, wherein said network
equipment comprises said first base station or management network apparatus which
manages said first base station and/or said at least one second base station.
9. A device for selecting user terminal so as to enhance calibration for reciprocity
error between uplink and downlink in a network equipment in time division duplex
communication systems, comprising:
an obtaining means, for obtaining channel quality related information between a
plurality of first user terminals and a first base station, wherein, said channel quality
related information is used for indicating channel qualities between the plurality of
first user terminals and said first base station;
a selecting means, for selecting, at least one said first user terminal as calibration
user terminal for calibrating said reciprocity error between uplink and downlink,
according to said channel quality-related information.
10. A device according to claim 9, wherein when said first base station works at
single cell mode, said selecting means is further used for:
- selecting at least one candidate user terminal from said plurality of first user
terminals according to said channel quality related information, wherein the channel
quality between said at least one candidate user terminal and said first base station is
greater than a first predetermined threshold;
- judging whether the number of said at least one candidate user terminal is
greater than a predetermined number N:
- taking N said user terminal which is the top N user terminal as to the
channel quality with said first base station in said at least one candidate user terminal
as said calibration user terminal, when the number of said at least one candidate user
terminal is greater than N;
- taking said at least one candidate user terminal as said calibration user
terminal when the number of said at least one candidate user terminal is less than or
equal to N.
11. A device according to claim 9, wherein when said first base station and at
least one neighboring second base station constitute a coordinated base station cluster
and work coordinately, said obtaining means is further used for:
-obtaining at least one first user terminal and a plurality of second user terminals,
wherein the channel quality between said at least one first user terminal and said first
base station is greater than a second predetermined threshold and the channel qualities
between said plurality of second user terminals and said at least one second base
station are greater than said second predetermined threshold;
said selecting means is further used for:
-taking at least one user terminal among said at least one first user terminal
and said plurality of second user terminals as candidate user terminals, wherein the
channel qualities between said at least one user terminal and at least two base stations
are greater than said second predetermined threshold;
-taking a plurality of said candidate user terminals which make said
coordinated base station cluster connected as said calibration user terminal, wherein
said coordinated base station cluster is connected when there is no nonzero element in
5 = (5,j ) = B , wherein B denotes node adjacent matrix, B denotes B to the
power of k, Z denotes total number of base stations in said coordinated base station
cluster.
12. A device according to claim 11, wherein, said selecting means further
comprises:
a removing means, for removing at least one candidate user terminal which has
minimal average channel quality one after another according to average channel
quality of said candidate user terminals, wherein said average channel quality of
candidate user terminals comprises the average value of channel qualities between
said candidate user terminal and said at least two base stations, and the channel
qualities between said candidate user terminal and at least two base stations are both
greater than said second predetermined threshold;
a first judging means, for judging whether the rest of said plurality of candidate
user terminals can make said coordinated base station cluster connected;
said removing means and said first judging means repeat above-mentioned
operation till said rest of said plurality of candidate user terminals can't make said
coordinated base station cluster connected, when said rest of said plurality of
candidate user terminals can make said coordinated base station cluster connected;
said selecting means is further used for taking the minimal number of said rest of
candidate user terminals which can make said coordinated base station cluster
connected as said calibration user terminal.
13. A device according to any one of claims 9 to 12, wherein when said channel
quality comprises signal to interference plus noise ratio information, said obtaining
means is further used for:
-obtaining downlink channel quality indication fed back by said plurality of first
user terminals;
-obtaining said signal to interference plus noise ratio information according to
said downlink channel quality indication.
14. A device according to claim 10, wherein said channel quality comprises
uplink channel estimation information obtained by said base station;
said obtaining means is further used for:
obtaining downlink channel quality information fed back by said calibration user
terminal;
said device further comprises:
a second judging means, for judging whether said downlink channel quality fed
back by said calibration user terminal is greater than said first predetermined
threshold;
said device is further used for calibrating said reciprocity error between uplink
and downlink according to said calibration user terminal, when said downlink channel
quality fed back by said calibration user terminal is greater than said first
predetermined threshold.
15. A device according to claim 11, wherein said channel quality comprises
uplink channel estimation information obtained by said base station;
said obtaining means is further used for:
obtaining downlink channel quality information fed back by said calibration user
terminal;
said device further comprises:
a third judging means, for judging whether said downlink channel quality fed
back by said calibration user terminal is greater than said second predetermined
threshold;
said device is further used for calibrating reciprocity error between uplink and
downlink according to said calibration user terminal, when said downlink channel
quality fed back by said calibration user terminal is greater than said second
predetermined threshold.