WO2016090611A1

WO2016090611A1 - Base station, mobile station and method thereof

Info

Publication number: WO2016090611A1
Application number: PCT/CN2014/093614
Authority: WO
Inventors: Bruno Clerckx; Mingbo DAI; Jaehyun Park; David Mazzarese; Kunpeng Liu
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2014-12-11
Filing date: 2014-12-11
Publication date: 2016-06-16
Also published as: CN105917594A

Abstract

The embodiment of the invention discloses a base station, a mobile station and a method thereof. The communication method for a mobile station which is served by a base station having at least one transmission point each equipped with multiple transmit antennas includes determining a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point; and reporting to the base station quantized channel state information CSI for the transmit antennas in the best subset. The embodiments of the invention can reduce the feedback overhead.

Description

BASE STATION, MOBILE STATION AND METHOD THEREOF

FIELD OF TECHNOLOGY

The invention relates to the technical field of the mobile communication, and in particular to a base station, a mobile station and a method thereof.

BACKGROUND OF THE INVENTION

Multiple antenna system such as Massive MIMO (Multi-Input Multi-Output) is a promising technology in the future communication system, such as 5G, for its ability to offer enormous enhancements in spectral efficiency and simple TX/RX (transmit/receive) structure. Ideally, more TX antennas enables larger benefit from antenna selection based on perfect CSIT (Channel State Information at Transmitter) , which incurs heavy feedback overhead and cost. Moreover, dedicated RF elements, such as A/D (Analog to Digital) convertor, D/A (Digital to Analog) convertor and amplifier, for each antenna incur enormous cost.

In a multiple antenna system with a limited number of RF (Radio Frequency) chains (or called RF hardware) , resource for feedback is limited. Therefore, it is nontrivial to investigate a practical setup with limited feedback and RF chains.

SUMMARY OF THE INVENTION

The embodiment of the invention relates to a base station, a mobile station and a method thereof to reduce the feedback overhead.

In a first aspect, there is provided a communication method for a mobile station which is served by a base station having at least one transmission point each equipped with multiple transmit antennas, the method comprising: determining a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point； and reporting to the base station quantized channel state information CSI for the transmit antennas in the best subset.

In a first possible implementation form of the method according to the first aspect as such, the determining the best subset of transmit antennas for each transmission point comprises: obtaining the channel gains of the multiple transmit antennas of the transmission point； and selecting the transmit antennas having the largest channel gains as the best subset.

In a second possible implementation form of the method according to the first aspect as such or according to the any of the preceding implementation forms of the first aspect, the determining the best subset of transmit antennas for each transmission point comprises: obtaining the channel gains of the multiple transmit antennas of the transmission point； and selecting the transmit antennas having the channel gains larger than a threshold as the best subset.

In a third possible implementation form of the method according to the first aspect as such or according to the any of the preceding implementation forms of the first aspect, the method further comprises: reporting to the base station the indices of the transmit antennas in the best subset .

In a fourth possible implementation form of the method according to the first aspect as such or according to the any of the preceding implementation forms of the first aspect, the determining the best subset of transmit antennas for each transmission point comprises: determining the best subset based on indication information received from the base station, the indication information indicating indices of the transmit antennas in the best subset which are selected by the base station according to channel gains.

In a fifth possible implementation form of the method according to the first aspect as such or according to the any of the preceding implementation forms of the first aspect, the reporting to the base station quantized channel state information CSI for the transmit antennas in the best subset comprises: reporting the quantized CSI using a code book whose size depends on size of the best subset.

In a sixth possible implementation form of the method according to the first aspect as such or according to the any of the preceding implementation forms of the first aspect, the method further comprises: receiving size information from the base station for indicating the size of the best subset； or selecting the size of the best subset at the mobile station.

In a seventh possible implementation form of the method according to the first aspect as such or according to the any of the preceding implementation forms of the first aspect, the method further comprises: reporting to the base station quantized CSI for the transmit antennas other than those in the best subset.

In a second aspect, there is provided a communication method for a base station having at least one transmission point each equipped with multiple transmit antennas, the method comprising: receiving a report from a mobile station for reporting quantized channel state information CSI for a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point； and transmitting data based on the quantized CSI.

In a first possible implementation form of the method according to the second aspect as such, the method further comprises: obtaining the channel gains of the multiple transmit antennas of the transmission point； and selecting the transmit antennas having the largest channel gains as the best subset.

In a second possible implementation form of the method according to the second aspect as such or according to the any of the preceding implementation forms of the second aspect, the method further comprises: obtaining the channel gains of the multiple transmit antennas of the transmission point； and selecting the transmit antennas having the largest channel gains as the best subset.

In a third possible implementation form of the method according to the second aspect as such or according to the any of the preceding implementation forms of the second aspect, the method further includes transmitting indication information to the mobile station for indicating to the mobile station the indices of the transmit antennas in the best subset.

In a fourth possible implementation form of the method according to the second aspect as such or according to the any of the preceding implementation forms of the second aspect, the method further includes receiving indication information from the mobile station for indicating the indices of the transmit antennas in the best subset which are selected by the mobile station according to channel gains.

In a fifth possible implementation form of the method according to the second aspect as such or according to the any of the preceding implementation forms of the second aspect, the method further includes the report of the quantized CSI uses a code book whose size depends on size of the best subset.

In a sixth possible implementation form of the method according to the second aspect as such or according to the any of the preceding implementation forms of the second aspect, receiving size information from the mobile station for indicating the size of the best subset； or selecting the size of the best subset at the base station.

In a seventh possible implementation form of the method according to the second aspect as such or according to the any of the preceding implementation forms of the second aspect, the method further includes receiving a report from the mobile station for reporting quantized CSI for the transmit antennas other than those in the best subset.

In a third aspect, there is provided mobile station which is served by a base station having at least one transmission point each equipped with multiple transmit antennas, the mobile station comprising: a determining unit configured to determine a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point； and a reporting unit configured to report to the base station quantized channel state information CSI for the transmit antennas in the best subset.

In a first possible implementation form of the mobile station according to the third aspect as such, the determining unit is configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.

In a second possible implementation form of the mobile station according to the third aspect as such or according to the any of the preceding implementation forms of the third aspect, the determining unit is configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.

In a third possible implementation form of the mobile station according to the third aspect as such or according to the any of the preceding implementation forms of the third aspect, the reporting unit is further configured to report to the base station the indices of the transmit antennas in the best subset.

In a fourth possible implementation form of the mobile station according to the third aspect as such or according to the any of the preceding implementation forms of the third aspect, the determining unit is configured to determine the best subset based on indication information received from the base station, the indication information indicating indices of the transmit antennas in the best subset which are selected by the base station according to channel gains.

In a fifth possible implementation form of the mobile station according to the third aspect as such or according to the any of the preceding implementation forms of the third aspect, the reporting unit is configured to report the quantized CSI using a code book whose size depends on size of the best subset.

In a sixth possible implementation form of the mobile station according to the third aspect as such or according to the any of the preceding implementation forms of the third aspect, the determining unit is further configured to determine the size of the best subset based on size information received from the base station, or to select the size of the best subset at the mobile station.

In a seventh possible implementation form of the mobile station according to the third aspect as such or according to the any of the preceding implementation forms of the third aspect, the reporting unit is further configured to report to the base station quantized CSI for the transmit antennas other than those in the best subset.

In a fourth aspect, there is provided a base station having at least one transmission point each equipped with multiple transmit antennas, the base station comprising: a receiver configured to receive a report from a mobile station for reporting quantized channel state information CSI for a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point； and a transmitter configured to transmitting data based on the quantized CSI.

In a first possible implementation form of the base station according to the fourth aspect as such, the base station further includes a processor configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.

In a second possible implementation form of the base station according to the fourth aspect as such or according to the any of the preceding implementation forms of the fourth aspect, the base station further includes a processor configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.

In a third possible implementation form of the base station according to the fourth aspect as such or according to the any of the preceding implementation forms of the fourth aspect, the transmitter is further configured to transmit indication information to the mobile station for indicating to the mobile station the indices of the transmit antennas in the best subset.

In a fourth possible implementation form of the base station according to the fourth aspect as such or according to the any of the preceding implementation forms of the fourth aspect, the receiver is further configured to receive indication information from the mobile station for indicating the indices of the transmit antennas in the best subset which are selected by the mobile station according to channel gains.

In a fifth possible implementation form of the base station according to the fourth aspect as such or according to the any of the preceding implementation forms of the fourth aspect, the report of the quantized CSI uses a code book whose size depends on size of the best subset.

In a sixth possible implementation form of the base station according to the fourth aspect as such or according to the any of the preceding implementation forms of the fourth aspect, the receiver is further configured to receive size information from the mobile station for indicating the size of the best subset； or the base station further comprises a processor configured to select the size of the best subset at the base station.

In a seventh possible implementation form of the base station according to the fourth aspect as such or according to the any of the preceding implementation forms of the fourth aspect, the receiver is further configured to receive a report from the mobile station for reporting quantized CSI for the transmit antennas other than those in the best subset.

Therefore the embodiments of the invention select a best subset of transmit antennas based on channel gains and perform feedback for the selected subset, thus reducing the feedback overhead.

BRIFE DESCRIPTION OF THE DRAWINGS

In order to more clearly describe the embodiments of the invention or the technical solutions in the prior art, accompanying drawings necessary for describing the embodiments or the prior art would be briefly described below. It is obvious to ordinary ones skilled in the art that these drawings described below are only for some embodiments of the invention and other drawings can be obtained from these drawings without any creative effort.

Fig. 1 shows an example of a massive MIMO system where the invention can be applied.

Fig. 2 shows a method of an embodiment of the invention.

Fig. 3 shows an example of gain from antennas selection according to an embodiment of the invention.

Fig. 4 shows an example of gain from antennas selection according to another embodiment of the invention.

Fig. 5 shows an example of gain from antennas selection according to another embodiment of the invention.

Fig. 6 shows an example of gain from antennas selection according to another embodiment of the invention.

Fig. 7 shows a method of another embodiment of the invention.

Fig. 8 shows a block diagram of a mobile station in an embodiment of the invention.

Fig. 9 shows a block diagram of a mobile station in another embodiment of the invention.

Fig. 10 shows a block diagram of a base station in an embodiment of the invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS

Hereinafter, the technical solutions in the embodiments of the invention would be clearly and fully described in connection with the accompanying drawings showing these embodiments. Obviously, the embodiments described are only a part of the embodiments of the invention, but not all the embodiments. Based on these embodiments, all other embodiments conceived by ordinary ones skilled in the art without any creative effort would fall within the protection scope of the invention.

The sum capacity of the system can be significantly improved by switching a subset of antennas to the RF chains. It has been shown that with more available antennas than RF chains, the antenna subset selection gain can be larger. Efficient searching algorithm has been proposed to select antennas, leading a nearly optimal performance with exhaustive search. However, this searching algorithm assumes perfect CSIT which is infeasible especially in massive MIMO system, where feedback overhead is too heavy to afford. On the other hand, with limited feedback, more transmit antennas does not lead to better sum rate performance due to the poor quantization quality. As the number of transmit antenna increases, the quantized CSI quality becomes lower and multiuser interference constrains the system performance.

Fig. 1 shows an example of a massive MIMO system where the invention can be applied, including a BS (base station) serving K MS (mobile station) : MS 1 through MS K.

The BS is equipped with large-scale antennas (N_T) and a limited number of RF chains (N_RF) , i.e., N_T >＝ N_RF. To fully understand how the system performance is related to the number RF chains, two different cases (N_T >N_RF or N_T ＝ N_RF) will be considered. A subset of transmit antennas is selected and switched to the RF chains in order to maximize the system (rate) performance. The selection procedure may be taken at the receiver side and the corresponding CSI is reported by limited backhaul resource. We assume an independent and identical distributed (i.i.d. ) Rayleigh channels.

For SU MISO (Single User, Multi-Input Single-Output) case, the optimal beamforming strategy to maximize the rate is matched beamforming (MBF) which requires accurate CSIT. However, reporting all CSI with limited feedback leads to large quantization loss. On the other hand, higher accuracy can be achieved by feeding back partial CSI only. Naturally, the idea is to select and report the strongest channels (with largest magnitudes) . Since the beamforming gain is proportional to the number of selected antennas, it is nontrivial to strike an optimal trade-off between the quantization quality and beamforming gain. Indeed, this method can be interpreted as a hybrid selection/MBF with limited feedback.

For MU MISO (Multiple Users, Multi-Input Single-Output) case, each MS reports the most significant CSI (with largest channel gains) and the transmitter regards the unreported ones as zero. A zero-forcing (ZF) beamforming strategy is considered based on the partial channel information. With massive transmit antennas, there is a high probability that MSs report mutually non-overlapping antenna subset. Then, the ZF beamformer which aims to cancel multiuser interference is also the matched beamforming for the desired signal. Indeed, we have residual multiuser interference due to the imperfectness of the reported CSI as well as the unreported CSI. However, the first term can be overcome if the size of the selected antenna subset is affordable with the available feedback bits. The second factor can be ignored due to the small channel gain of the unreported channels. To sum up, the multiplexing gain can be achieved while we have gain over the conventional MU MISO system.

Moreover, the system performance can be further improved if we can eliminate the antenna selection overhead. Since the MS needs to inform the BS which antenna subset is selected, the resulted overhead can be very heavy in massive MIMO system. In the literature, has investigated the predictability of up (down) -link from the down (up) -link in FDD system. It implies that, to some extent, the channel ordering information (COI) is known for the BS. Under this model, all the feedback bits can be used to quantize the channels. To reveal the potential gain, we will investigate the cases with/without antenna selection overhead.

Perfect CSI at the receiver (CSIR) is assumed and a number of ‘B’ feedback bits are available at each MS. At each time instant, the receiver of MS first select the strongest channels with largest magnitudes based on instantaneous CSI and report the index of the selection combination and the corresponding CSI to the transmitter of BS with random vector quantization (RVQ) . A conventional/reference scheme is defined as N_T ＝ N_RF with ‘B’ feedback bits. The invention tries to investigate when the proposed scheme outperforms the reference scheme and how the gain scales with NT and B. Namely, the invention addresses how to maximally extract antenna selection gain under a practical setup (limited feedback and RF chains) by dynamic antenna selection.

Fig. 2 shows a communication method of an embodiment of the invention. The method of Fig. 2 is performed by a MS, such as one of MS 1 to MS K as shown in Fig. 1, which is served by a BS having at least one transmission point each equipped with multiple transmit antennas.

201, the MS determines a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point.

A BS may include one or more transmission point, such as antenna array, antenna group, DA (distributed antenna) elements, etc. . Each transmission point is equipped with multiple (N_T) transmit antenna as shown in Fig. 1 where N_T may vary or be fixed for different transmission points.

202, the MS reports to the BS quantized CSI for the transmit antennas in the best subset.

In particular, the best subset has the most significant CSI among all the transmit antennas. The CSI of other transmit antennas than those in the best subset may not be reported so that the feedback overhead is reduced while the negative impact due to such imperfect CSI report can be controlled to an acceptable extent.

The subset selection may be performed at the MS. As a proposed non-limitative scheme, in step 201, the MS may obtain the channel gains of all the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.

This proposed scheme is compared with a reference scheme where N_RF out of N_T antennas are randomly selected to transmit hereinafter.

SU case 1: without RF chain constraint and with selection overhead

Consider the proposed scheme where M out of N_T antennas is selected, the average output SNR is given by:

where

and the input SNR ＝ P/σ²is enhanced by beamforming and attenuated by quantization. Beamforming gain consists of two terms: the former denotes array gain and the latter represents antenna selection gain. To inform the transmitter (BS) which antennas are selected, the receiver (MS) consumes certain bits (selection overhead) to send the index of the antenna subset while the effective bits are used to quantize the selected CSI.

The average output SNR of the reference scheme by using N_T antennas:

It can be shown that (1) <＝ (3) , implying the proposed scheme does not provide gain in this case. Using all N_T antennas is a better choice.

SU case 2: with RF chain constraint and with selection overhead

Consider the proposed scheme where N_RF out of N_T antennas are selected, the average output SNR is given by:

where

While the average output SNR of the reference scheme by randomly selecting N_RF out of N_T antennas is given by

Note that (6) is a monotonically increasing function with respect to N_RF and finally saturated at a fixed point below:

which is only a function of input SNR and B. It implies that with limited feedback, we prefer to utilize as many RF chains as available. However, the marginal gain from the RF chains increase is get smaller and smaller. On the other hand, for fixed the number of RF chains, the criteria of B and N_T by which the proposed scheme has gain over the reference scheme is identified and also how the gain changes with B and N_T is investigated.

a) Fix N_RF and N_T

With the help of harmonic series and binomial coefficient,

To achieve some gain, the number of feedback bits should satisfy (4) . The gain ΔSNR ＝ SNR_cand-SNR_ref scales with

ΔSNR～ (α-2^-γB) (9)

where α, γ are constant independent of B. When B goes to infinity,

For example, N_T/N_RF＝50, ΔSNR_dB＝7dB. The gain scales with log (N_T) .

b) Fix N_RF and B

Note that (8) is a sufficient condition for (7) . Then, we obtain the criteria of N_T

which implies that the ratio is bounded, namely, N_T/N_RF≤ρ (B, N_RF) in order to achieve gain from the proposed scheme with limited B. As N_T increases, the selection overhead approaches/exceeds the number of available bits and the number of effective/residual bits is too small to accurately quantize the channels. In contrast to the prior art, increase the number of TX antennas does not always benefit us. When B is small or N_T is large, the proposed scheme performs worse than the reference scheme. We may have gain by increasing B or reducing N_T.

c) Fix N_RF, N_T and B

In this case, the proposed scheme may perform worse than the reference scheme. Gain can be obtained by using partial RF chains. By doing so, the selection overhead is reduced and more feedback bits give rise to higher quantization accuracy. Meanwhile, the transmit array gain becomes lower. An optimization procedure should be taken to achieve the best balance between the benefits and drawbacks of partial RF usage.

Denote M as number of RF chains used and find the optimum M <＝ N_RF to maximize the average output SNR and feedback the corresponding CSI subset. A general optimization function is defined as follows

which requires a complex exhaustive search algorithm to achieve the optimal solution. Fortunately, a low complexity selection method to maximize the SNR exists in our case. More specifically, sort the channels by magnitude and report the largest ones.

The average output SNR in this case is given by

Due to the coupled property, it is difficult to obtain a closed-form solution of M by which the proposed scheme overcomes the reference scheme. A numerical optimization procedure over 1< M <＝ N_RF is required. It implies that the proposed scheme provides at least a lower bound on the gain. When B is large, exploiting all RF chains is optimal and (16) boils down to (4) . Otherwise, numerically optimize (16) and compare it with the reference scheme, determine to increase B or decrease N_T.

Suppose the transmitter has the knowledge of channel ordering information (COI) in advance, the proposed scheme can be further enhanced. For FDD system, we can predict uplink channel from downlink channel or vice versa. Briefly, a finite-order autoregressive (AR) process can be used to model the frequency selective fading (energy) on up/downlink. When the relative difference in the up/downlink carrier frequency is small, they show that both links share the common AR coefficients. Therefore, both links experience similar variation trends. To this end, the assumption COI at the transmitter is valid and the selection overhead can be eliminated. To reveal the potential gain, we reconsider the two cases above without selection overhead.

SU case 3: without RF chain constraint and without selection overhead

Suppose M antennas are selected, the average output SNR of the proposed scheme:

while the reference scheme using all N_T antennas:

It is difficult to derive the closed-form expression of M by which the proposed scheme overcomes the reference, i.e., (17) > (18) . However, it is easy to show by Monte-Carlo simulation that by selecting partial transmit antennas, we can achieve gain. The gain depends on NT, NRF, and B. An example is given by Fig. 3 where there is a lower bound on which the proposed scheme outperforms the reference. The maximum gain is achieved in the middle region by trading off beamforming gain with CSI accuracy.

SU case 4: with RF chain constraint and without selection overhead

Suppose M antennas are selected, the average output SNR of the proposed scheme:

while the reference scheme randomly selecting N_RF out of N_T antennas:

More gain than SU case 3 can be obtained since the performance of the reference scheme becomes poor due to limited RF chains, which is shown in Fig. 3 where the curve “Pro” and “Ref” corresponds to the proposed scheme and the reference scheme, respectively.

Next, consider a multiuser MISO Broadcast Channel (BC) , zero-forcing beamforming strategy is adopted to extract multiplexing gain. Residual multiuser interference occurs due to imperfect reported CSI as well as unreported CSI. The proposed scheme is the same as the SU case described above and two reference schemes are present.

The proposed scheme: sort the channel gain and select the largest M, and report the index of the selected subset and the corresponding CSI for that subset

The reference scheme 1: fix number of MS as Nu and report CSI of the first M antennas (equivalently, randomly select M out of N_T antennas) to transmit

The reference scheme 2: fix number of MS equal to the number of reported antennas (Nu ＝ M) and report CSI of the first M antennas (equivalently, randomly select M out of N_T antennas) to transmit

MU case 1: without RF chain constraint and with selection overhead

The average rate of the proposed scheme is given by

where the expectation is taken over the channel as well as the RVQ codebook. v_max, _i denotes the beamforming vector computed with partial (largest channel gain) CSI by viewing unreported CSI as zero. Meanwhile, the average rates of the two reference scheme are written as

The closed-form approximations of (22) and (23) are derived later while the closed-form expression is still open. However, by simulation in Fig. 4 (as shown by curves 401 to 403) , we can observe that the proposed scheme fails to provide gain over the references due to the heavy selection overhead, especially in the large-scale transmit antenna system. Again, we can eliminate this overhead with COI at the transmitter.

MU case 2: without RF chain constraint and without selection overhead

The proposed scheme in this case indeed achieves gains over the two reference schemes since

1) it ignores the selection overhead and obtain the same CSI quality as

reference schemes

1 and 2；

2) it keeps most of channel information by reporting the most significant channels； and

3) if MSs select mutually non-overlapping subset, a simple matched beamforming on the reported subset is also ZFBF for other MSs； multiuser interference is small since it originates from unreported channels (with relatively small channel gain) .

Fig. 5 indicates the simulation result from which the proposed scheme obtains gain over the two references.

It can be seen from curves 501 to 503 that the maximum gain is achieved with mediate M. The reason is following. For small M, CSI is quantized with high quality while non-overlapping subset with high probability leads to high beamforming gain. However, it corresponds to low multiplexing gain and large multiuser interference due to many unreported CSI. On the other side, large M (approaches N_T) corresponds to the other way around. Therefore, mediate M achieves the optimal balance among those factors.

Note that the gain largely depends on the non-overlapping subset which facilitates the ZF and matched beamforming simultaneously. The number of reported partial CSI must be comparable to N_T, otherwise the multiuser interference degrades the system performance. On the other hand, if we fix the number of Nu and increase the number of Nt, we can finally get non-overlapping subset. But the multiuser interference becomes dominant due to the fact that the number of unreported CSI is larger.

However, as N_T increases and if the transmitter can dynamically decide the number of MSs, the optimal number of MSs will also be increased. By doing this, the inter-user interference can be small (we feedback sufficient number of significant CSI) . Then, we equivalently have Nu parallel SU-MISO case. For each MS, we have shown that the marginal gain is small when N_T becomes larger. This is what we observe from Fig. 6.

When the number of MS is very large, the proposed scheme can be easily combined with multiuser diversity, i.e., MSs with non-overlapping subset will be selected to transmit. The number of MSs and the number of reported CSI for each MS should be determined by the transmitter based on the closed-form/approximation of sum rate. Given N_T, B, this can be done by maximizing the sum rate expression which is function of N_T, B and M. MS selection algorithm is also required.

Though it is shown by simulation that the proposed scheme has gain over the references, analytical results would be helpful to get insight on how these system parameters affect the rate performance. Here we present the closed-form approximations of (22) and (23) and provide a guideline to compute the closed-form expression for (21) .

To calculate the expectation in the numerator and denominator of (26) , we need to build probability model to characterize the average beamforming gain (mainly determined by matched beamforming gain from non-overlapping subset)

MU case 3: with RF chain constraint and without selection overhead

In this case, how to switch RF chains to partial transmit antenna is very important. RF switch needs to be optimized at the transmitter by exhaustive search. Note that MU case 3 gives a lower bound of MU case 2. The simulation result shows that the proposed scheme performs worse than the MU case 2, however, still have gain over the reference scheme.

Alternatively, as another embodiment, instead of or in addition to selection based on sorting of gains, the subset selection may be based on a threshold which may be decided and configured by a controller of the communication system, such as control functionality in the base station.

In particular, in step 201, the MS may obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.

Optionally as another embodiment, CSI for the transmit antennas other than those in the best subset may further be reported to the BS. That is, some or all of the transmit antennas other than those in the best subset may compose a worse subset and the MS may report the CSI for the worse subset. The selection for the worse subset may be similar to that for the best subset, for example, based on sorting of the channel gains and select the smallest ones, or based on another threshold for selecting the transmit antennas with channel gains lower than this threshold. As another embodiment, the worse subset may be simply the remaining antennas after selecting the best subset.

Alternatively, the subset selection may be performed at the BS side, in a similar way to that performed at the MS side as described above, in which case, in step 201, the MS may determine the best subset based on indication information received from the base station, the indication information indicating indices of the transmit antennas in the best subset which are selected by the base station according to channel gains.

Further, when the MS determine the best subset and/or the worse subset, the MS may report to the BS the indices of the transmit antennas in the subset (s) . On the other hand, when the BS determine the best subset and/or the worse subset, the BS may indicate to the MS the indices of the transmit antennas in the subset (s)

Optionally, the MS may report the quantized CSI using a code book whose size depends on size of the best subset.

The size of the best subset may be decided by the BS, in which case the MS may receive size information from the base station for indicating the size of the best subset. Or the size of the best subset may be decided by the MS per se, and for example, the MS may select the size of the best subset.

In summary, the MS may obey to the following behavior:

1) If the BS or transmission point has a number of RF chains equal to the number of transmit antennas, the MS reports its quantized CSI for all transmit antennas, i. e. the best subset is equal to the whole set of available antennas.

2) If the BS or transmission point has a number of RF chains lower than the number of transmit antennas and if the MS is requested to report an antenna subset whose size is equal to the number of RF chains, the MS selects and reports the best antenna subset and the quantized CSI of that subset only if the number of feedback bits available is larger than a certain threshold (being a function of the number of transmit antennas and the number of RF chains) , otherwise the MS reports the quantized CSI for a predefined set of antennas, whose number is equal to the number of RF chains.

3) If the BS or transmission point has a number of RF chains lower than the number of transmit antennas and if the MS is requested to report an antenna subset whose size is equal to the number of RF chains, for a given number of feedback bits, the MS reports the best antenna subset and the quantized CSI of that subset only if the number of transmit antennas is smaller than a certain threshold (being a function of the number of the number of feedback bits and the number of RF chains) , otherwise the MS reports the quantized CSI for a predefined set of antennas, whose number is equal to the number of RF chains.

4) If the BS or transmission point has a number of RF chains lower than the number of transmit antennas and if the MS is not restricted to report any specific antenna subset size, the MS selects the best antenna subset and reports its index and the corresponding quantized CSI. In this case, the antenna subset size can be strictly smaller than the number of RF chains.

To sum up, with limited feedback and/or RF chains constraint, increasing the number of TX antennas can benefit in SU/MU MISO scenarios by dynamic and user-specific antenna selection.

Fig. 7 shows a communication method of another embodiment of the invention. The method of Fig. 7 is performed by a BS, such as BS as shown in Fig. 1, which has at least one transmission point each equipped with multiple transmit antennas.

701, the BS receives a report from a MS for reporting quantized channel state information CSI for a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point.

702, the BS transmits data based on the quantized CSI.

The processes of method in Fig. 7 may be corresponding to or similar to the processes of method in Fig. 2, and therefore will not be described in detail.

Optionally, the subset selection may be performed at the BS in which case in step 701 the BS may obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.

Alternatively, in step 701 the BS may obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.

Alternatively, as another embodiment, the subset selection may be performed at the MS in which case in step 701 the BS may receive indication information from the mobile station for indicating the indices of the transmit antennas in the best subset which are selected by the mobile station according to channel gains.

Further, when the MS determine the best subset and/or the worse subset, the MS may report to the BS the indices of the transmit antennas in the subset (s) . On the other hand, when the BS determine the best subset and/or the worse subset, the BS may indicate to the MS the indices of the transmit antennas in the subset (s) by for example transmitting corresponding indication information to the MS.

Fig. 8 shows a block diagram of a mobile station in an embodiment of the invention. The mobile station 80 is served by a base station having at least one transmission point each equipped with multiple transmit antennas. As shown in Fig. 8, the mobile station includes a determining unit 81 and a reporting unit 82.

The determining unit 81 is configured to determine a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point.

The reporting unit 82 is configured to report to the base station quantized channel state information CSI for the transmit antennas in the best subset.

The parts of the mobile station 80 can perform corresponding process relating to MS as described in the embodiments based on Fig. 1 to Fig. 6 and will not be repeated redundantly herein for the propose of simplicity.

Optionally, as an embodiment, the determining unit 81 is configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.

Optionally, as another embodiment, the determining unit 81 is configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.

Optionally, as another embodiment, the reporting unit 82 is further configured to report to the base station the indices of the transmit antennas in the best subset.

Optionally, as another embodiment, the determining unit 81 is configured to determine the best subset based on indication information received from the base station, the indication information indicating indices of the transmit antennas in the best subset which are selected by the base station according to channel gains.

Optionally, as another embodiment, the reporting unit 82 is configured to report the quantized CSI using a code book whose size depends on size of the best subset.

Optionally, as another embodiment, the determining unit 81 is further configured to determine the size of the best subset based on size information received from the base station, or to select the size of the best subset at the mobile station.

Optionally, as another embodiment, the reporting unit 82 is further configured to report to the base station quantized CSI for the transmit antennas other than those in the best subset.

As shown in Fig. 9 the mobile station 90 include a processor 91, a transmitter 92 and a receiver 93.

The processor 91 is configured to determine a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point.

The transmitter 92 is configured to report to the base station quantized channel state information CSI for the transmit antennas in the best subset.

The parts of the mobile station 90 can perform corresponding process relating to MS as described in the embodiments based on Fig. 1 to Fig. 6 and will not be repeated redundantly herein for the propose of simplicity.

Optionally, as an embodiment, the processor 91 is configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.

Optionally, as another embodiment, the processor 91 is configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.

Optionally, as another embodiment, the transmitter 92 is further configured to report to the base station the indices of the transmit antennas in the best subset.

Optionally, as another embodiment, the processor 91 is configured to determine the best subset based on indication information received from the base station by the receiver 93, the indication information indicating indices of the transmit antennas in the best subset which are selected by the base station according to channel gains.

Optionally, as another embodiment, the transmitter 92 is configured to report the quantized CSI using a code book whose size depends on size of the best subset.

Optionally, as another embodiment, the processor 91 is further configured to determine the size of the best subset based on size information received from the base station, or to select the size of the best subset at the mobile station.

Optionally, as another embodiment, the transmitter 92 is further configured to report to the base station quantized CSI for the transmit antennas other than those in the best subset.

Fig. 10 shows a block diagram of a base station in an embodiment of the invention. The base station 100 may have at least one transmission point each equipped with multiple transmit antennas. As shown in Fig. 10 the base station 100 includes a receiver 101 and a transmitter 102.

The receiver 101 is configured to receive a report from a mobile station for reporting quantized channel state information CSI for a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point.

The transmitter 102 configured to transmitting data based on the quantized CSI.

The parts of the base station 100 can perform corresponding process relating to BS as described in the embodiments based on Fig. 1 to Fig. 6 and will not be repeated redundantly herein for the propose of simplicity.

Optionally, as an embodiment, the base station 100 may further include a processor 103 configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.

Optionally, as another embodiment, the base station 100 may further include a processor 103 configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.

Optionally, as another embodiment, the transmitter 102 is further configured to transmit indication information to the mobile station for indicating to the mobile station the indices of the transmit antennas in the best subset.

Optionally, as another embodiment, the receiver 101 is further configured to receive indication information from the mobile station for indicating the indices of the transmit antennas in the best subset which are selected by the mobile station according to channel gains.

Optionally, as another embodiment, the report of the quantized CSI uses a code book whose size depends on size of the best subset.

Optionally, as another embodiment, the receiver 101 is further configured to receive size information from the mobile station for indicating the size of the best subset.

Optionally, as another embodiment, the base station 100 may further include a processor 103 configured to select the size of the best subset at the base station.

Optionally, as another embodiment, the receiver 101 is further configured to receive a report from the mobile station for reporting quantized CSI for the transmit antennas other than those in the best subset.

In the embodiments of the invention, the MS may be any one as follows, may be stationary or mobile, and the example of the stationary MS may include user equipment, terminal, mobile station, subscriber unit or station, etc. The example of the mobile MS may include cellular phone, personal digital assistant (PDA) , wireless modem, wireless communication device, handheld device, laptop computer, cordless phone or wireless local loop (WLL) station, etc.

It is to be noted that the term such as “first, second, and so on” used in the context is only used to distinguish one entity or operation from another entity or operation, and is not intended to the actual relationship or sequence between these entities or operations. Furthermore, the term “include” , “comprise” or the other variation is intended to a inclusion in inclusive sense, but not in exclude sense, thus the process, method, object or equipment including elements will not only include these elements, but also include other elements which are not clearly listed, or further include the elements inherent to the process, method, object or equipment. Unless defined otherwise, the elements defined by the term “comprise a …” would not exclude the presence of other identical elements in the process, method, object or equipment including the stated elements.

From the description of the embodiment of the invention, ones skilled in the art would clearly understand that the invention can be achieved by software together with the necessary general-purpose hardware, and certainly can also be achieved only by hardware, but the former would be preferred. Based on this understanding, the solution of the invention naturally or the portion by which the invention contributes to the prior art can be implemented in the form of software products, and the software products can be stored in storage media, such as ROM/RAM, hard disks, compact disks and the like, containing several instructions capable of enabling a computer device (personal computer, server or network device, etc. ) to execute the method described in the embodiments or part of the embodiments.

While the invention has been described by way of the preferred embodiments, it is to be noted that many modifications or variations can be made in the embodiments by ordinary ones skilled in the art as these modifications or variations should also fall within the protection scope of the invention.

Claims

A communication method for a mobile station which is served by a base station having at least one transmission point each equipped with multiple transmit antennas, the method comprising:

determining a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point； and

reporting to the base station quantized channel state information CSI for the transmit antennas in the best subset.
The method of claim 1, wherein the determining the best subset of transmit antennas for each transmission point comprises:

obtaining the channel gains of the multiple transmit antennas of the transmission point； and

selecting the transmit antennas having the largest channel gains as the best subset.
The method of claim 1, wherein the determining the best subset of transmit antennas for each transmission point comprises:

obtaining the channel gains of the multiple transmit antennas of the transmission point； and

selecting the transmit antennas having the channel gains larger than a threshold as the best subset.
The method of claim 2 or 3, further comprising:

reporting to the base station the indices of the transmit antennas in the best subset.
The method of claim 1, wherein the determining the best subset of transmit antennas for each transmission point comprises:

determining the best subset based on indication information received from the base station, the indication information indicating indices of the transmit antennas in the best subset which are selected by the base station according to channel gains.
The method of any one of claims 1 to 5, wherein the reporting to the base station quantized channel state information CSI for the transmit antennas in the best subset comprises:

reporting the quantized CSI using a code book whose size depends on size of the best subset.
The method of claim 5, further comprising:

receiving size information from the base station for indicating the size of the best subset； or

selecting the size of the best subset at the mobile station.
The method of any one of claims 1 to 7, further comprising:

reporting to the base station quantized CSI for the transmit antennas other than those in the best subset.
A communication method for a base station having at least one transmission point each equipped with multiple transmit antennas, the method comprising:

receiving a report from a mobile station for reporting quantized channel state information CSI for a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point； and

transmitting data based on the quantized CSI.
The method of claim 9, further comprising:

obtaining the channel gains of the multiple transmit antennas of the transmission point； and

selecting the transmit antennas having the largest channel gains as the best subset.
The method of claim 9, further comprising:

obtaining the channel gains of the multiple transmit antennas of the transmission point； and

selecting the transmit antennas having the channel gains larger than a threshold as the best subset.
The method of claim 10 or 11, further comprising:

transmitting indication information to the mobile station for indicating to the mobile station the indices of the transmit antennas in the best subset.
The method of claim 9, further comprising:

receiving indication information from the mobile station for indicating the indices of the transmit antennas in the best subset which are selected by the mobile station according to channel gains.
The method of any one of claims 9 to 13, wherein the report of the quantized CSI uses a code book whose size depends on size of the best subset.
The method of claim 14, further comprising:

receiving size information from the mobile station for indicating the size of the best subset； or

selecting the size of the best subset at the base station.
The method of any one of claims 9 to 15, further comprising:

receiving a report from the mobile station for reporting quantized CSI for the transmit antennas other than those in the best subset.
A mobile station which is served by a base station having at least one transmission point each equipped with multiple transmit antennas, the mobile station comprising:

a determining unit configured to determine a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point； and

a reporting unit configured to report to the base station quantized channel state information CSI for the transmit antennas in the best subset.
The mobile station of claim 17, wherein the determining unit is configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.
The mobile station of claim 17, wherein the determining unit is configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.
The mobile station of claim 18 or 19, the reporting unit is further configured to report to the base station the indices of the transmit antennas in the best subset.
The mobile station of claim 17, wherein the determining unit is configured to determine the best subset based on indication information received from the base station, the indication information indicating indices of the transmit antennas in the best subset which are selected by the base station according to channel gains.
The mobile station of any one of claims 17 to 21, wherein the reporting unit is configured to report the quantized CSI using a code book whose size depends on size of the best subset.
The mobile station of claim 22, wherein the determining unit is further configured to determine the size of the best subset based on size information received from the base station, or to select the size of the best subset at the mobile station.
The mobile station of any one of claims 1 to 7, wherein the reporting unit is further configured to report to the base station quantized CSI for the transmit antennas other than those in the best subset.
A base station having at least one transmission point each equipped with multiple transmit antennas, the base station comprising:

a receiver configured to receive a report from a mobile station for reporting quantized channel state information CSI for a best subset of transmit antennas for each transmission point which is selected, based on channel gains, among the multiple transmit antennas of the transmission point； and

a transmitter configured to transmitting data based on the quantized CSI.
The base station of claim 25, further comprising a processor configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the largest channel gains as the best subset.
The base station of claim 25, further comprising a processor configured to obtain the channel gains of the multiple transmit antennas of the transmission point, and select the transmit antennas having the channel gains larger than a threshold as the best subset.
The base station of claim 26 or 27, wherein the transmitter is further configured to transmit indication information to the mobile station for indicating to the mobile station the indices of the transmit antennas in the best subset.
The base station of claim 25, wherein the receiver is further configured to receive indication information from the mobile station for indicating the indices of the transmit antennas in the best subset which are selected by the mobile station according to channel gains.
The base station of any one of claims 25 to 29, wherein the report of the quantized CSI uses a code book whose size depends on size of the best subset.
The base station of claim 30, wherein

the receiver is further configured to receive size information from the mobile station for indicating the size of the best subset； or

the base station further comprises a processor configured to select the size of the best subset at the base station.
The method of any one of claims 25 to 31, wherein the receiver is further configured to receive a report from the mobile station for reporting quantized CSI for the transmit antennas other than those in the best subset.