WO2006093468A1

WO2006093468A1 - Transmit power allocation in wireless communication system

Info

Publication number: WO2006093468A1
Application number: PCT/SG2005/000068
Authority: WO
Inventors: Choo Eng Yap; Ping Luo
Original assignee: Matsushita Electric Industrial Co., Ltd.
Priority date: 2005-03-04
Filing date: 2005-03-04
Publication date: 2006-09-08

Abstract

A method for transmit power allocation in wireless communication systems having a plurality of antenna branches in both the transmitter and the receiver and a plurality of sub-carriers implemented for each antenna branch, the method comprising providing a feedback signal based on measured link parameters of one or more of the respective antenna branches; determining sizes of sub-carrier groups for the antenna branches based on the feedback signal; determining numbers of the sub-carrier groups for the antenna branches based on the respective sub-carrier group sizes; and performing transmit power allocation based on the determined sub-carrier groups for the antenna branches.

Description

TRANSMIT POWER ALLOCATION IN WIRELESS COMMMUNICATION SYSTEM

FIELD OF INVENTION

The present invention relates broadly to a method for transmit power allocation in wireless communication systems having a plurality of antenna branches in both the transmitter and the receiver and a plurality of sub-carriers implemented for each transmit antenna branch, to a wireless communication system, to a transmitter for a wireless communication system, and to a receiver for a wireless communication system.

BACKGROUND

Simultaneous transmission of multiple data streams occurs in multiple-input multiple-output (MIMO) communication systems that employ multiple transmit antennas (N₁ ) and multiple receive antennas (N _r )- MIMO systems can help to improve performance through spatial diversity or increase system capacity by means of spatial multiplexing. Such improvements are made possible by reducing the effect of random fading and multi-path delay spread in a wireless communication system.

The multiple communication channels present between the transmit and receive antennas typically experience different link conditions which will vary with time. MIMO systems with feedback provide the transmitter with channel state information, allowing the use of methods such as link adaptation and waterfilling to yield a higher level of performance.

MIMO techniques were first designed assuming a narrowband wireless system, i.e. a flat fading channel. To overcome the frequency selective channels posed by the wireless environment, orthogonal frequency division multiplexing (OFDM) is used in conjunction with MIMO systems. Using the inverse fast Fourier transform (IFFT), OFDM is able to convert each frequency selective channel into a set of parallel frequency-flat sub-channels or sub- carriers. The frequencies of these sub-channels are orthogonal and overlapping to one another, hence improving spectral efficiency and also minimize inter-carrier interference. The addition of a cyclic prefix to the OFDM symbol further helps to reduce the multi-path effects.

The multiple communication channels present between transmit and receive antennas typically experience different link conditions, which will vary with time. MIMO systems with feedback provide the transmitter with the channel state information, allowing the use of link adaptation methods to achieve a higher system performance. Channel state information, such as sub-carrier channel Signal-to-Noise ratio (SNR) for the propagation path between each transmit-receive antenna pair, is usually used for adaptation. Transmit power allocation based on channel state information is effective in adapting power allocation to varying channel conditions. However, power allocation using waterfilling methods based on received sub-carrier channel SNR have a high computational complexity.

In N_t χ N_r MIMO OFDM system, antenna number and multilevel modulation (MLM) are factors that increase system throughput. However, system complexity increases when large numbers of antennas and high modulation levels are used.

For example, the water-filling transmit power allocation method for a MIMO OFDM system with large number of sub-carriers requires high computational processing and large feedback overheads. In response to the challenge of obtaining high throughput and low error rates, it is desired to reduce the complexity of information processing and to lower the overheads required during feedback of channel information in order to improve the efficiency for transmit power allocation in a MIMO OFDM system.

SUMMARY

In accordance with a first aspect of the present invention there is provided a method for transmit power allocation in wireless communication systems having a plurality of antenna branches in both the transmitter and the receiver and a plurality of sub-carriers implemented for each antenna branch, the method comprising providing a feedback signal based on measured link parameters of one or more of the respective antenna branches; determining sizes of sub-carrier groups for the antenna branches based on the feedback signal; determining numbers of the sub- carrier groups for the antenna branches based on the respective sub-carrier group sizes; and performing transmit power allocation based on the determined sub-carrier groups for the respective antenna branches.

The link parameter of each antenna branch may comprise one or more of a group consisting of variance, RMS delay, τ_rms, and coherence bandwidth over all sub- carriers of each antenna branch.

The feedback signal may be based on an extreme measured link parameter for all antenna branches, and one sub-carrier group size is determined for all antenna branches based on the extreme measured link parameter.

The feedback signal may be based on each measured link parameter for the respective antenna branches, and the group sizes for the respective antenna branches are individually determined.

The power allocation may be first performed across all antenna branches, and then across all sub-carrier groups and all sub-carriers of each sub-carrier group for each antenna branch.

The power allocation across all antenna branches may be based on a measured link property of the antenna branches.

The power allocation may be based on the equal power allocation law.

The power allocation may be based on the water-filling power allocation law.

The power allocation across all sub-carrier groups for each antenna branch may be based on mean channel gain values of the respective sub-carrier groups. The power allocation may be based on the equal power allocation law.

The power allocation may be based on the water-filling power allocation law.

The power allocation may comprise a threshold method.

Power may be equally allocated to all sub-carriers of each sub-carrier group for each antenna branch.

The power allocation may be first performed across all sub-carrier groups for all antenna branches, and then across all antenna branches for each sub-carrier group and all sub-carriers for each sub-carrier group for each antenna branch.

Power may be allocated equally to all sub-carrier groups for all antenna branches.

The power allocation for each sub-carrier group across all antenna branches may be performed based on threshold decisions or a water-filling method.

Power may be equally allocated across all sub-carriers for each sub-carrier group for each antenna branch.

In accordance with a second aspect of the present invention there is provided a wireless communication system comprising a transmitter; a receiver; a plurality of antenna branches implemented in both the transmitter and receiver, each antenna branch arranged for a plurality of sub-carriers; a measurement unit for measuring link parameters of the respective antenna branches; a feedback unit providing a feedback signal based on the measured link parameters; a group size setting module determining sizes of sub-carrier groups for the antenna branches based on the feedback signal; a group number setting module for determining numbers of the sub-carrier groups for the antenna branches based on the respective sub-carrier group sizes; and a transmit power allocation processor performing transmit power allocation based on the sub-carrier groups for the respective antenna branches. In accordance with a third aspect of the present invention there is provided a transmitter for a wireless communication system, the transmitter comprising a plurality of antenna branches implemented in the transmitter, each antenna branch arranged for a plurality of sub-carriers; a feedback unit for receiving a feedback signal based on measured link parameters of the respective antenna branches; a group size setting module determining sizes of sub-carrier groups for the antenna branches based on the feedback signal; a group number setting module for determining numbers of the sub-carrier groups for the antenna branches based on the respective sub-carrier group sizes; and a transmit power allocation processor performing transmit power allocation based on the sub-carrier groups for the respective antenna branches.

In accordance with a fourth aspect of the present invention there is provided a receiver for a wireless communication system, the receiver comprising a plurality of antenna branches implemented in the receiver, each antenna branch arranged for a plurality of sub-carriers; a measurement unit for measuring link parameters of the respective antenna branches; a feedback unit for providing a feedback signal based on the measured link parameters for determining sizes of sub-carrier groups for the antenna branches based on the feedback signal and for determining numbers of the sub-carrier groups for the antenna branches based on the respective sub-carrier group sizes.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will be better understood and readily apparent to one with ordinary skill in the art from the following written description, by way of example only, and in conjunction with the drawings, in which:

Figure 1 is the block diagram of the transmitter for use in a communication system in an example embodiment.

Figure 2 is the block diagram of the receiver for use in a communication system in an example embodiment. Figure 3 is the block diagram of a MIMO OFDM communication system with low complexity transmit power allocation method in an example embodiment.

Figure 4 is the flowchart of the algorithm for setting of sub-carrier group size based on channel variance in an example embodiment.

Figure 5 is the flowchart of spatial-frequency transmit power allocation algorithm in an example embodiment.

Figure 6 is the flowchart of frequency-spatial transmit power allocation algorithm in an example embodiment.

Figure 7 shows a flow-chart illustrating a method for transmit power allocation in wireless communication systems having a plurality of antenna branches and a plurality of sub-carriers on each antenna branch, in accordance with an example embodiment.

DETAILED DESCRIPTION

Figure 1 shows a transmitter 100 for a communication system in an example embodiment. Figure 2 shows a receiver 200 for a communication system in an example embodiment. Although both Figures 1 and 2 show the communication system employing only two transmit and receive antennas respectively, the number of transmit and receive antennas are not restricted. The system 100, 200 is able to employ multiple transmit antennas ( N_r ) and multiple receive antennas ( N_r ).

Turning initially to Figure 1 , at the transmitter 100, separate data processing is performed for each individual antenna 122, 122'. In other words, different streams of independent data are transmitted from each of the transmit antennas 122, 122'.

Describing then one such stream, the binary data input 101 firstly passes through a cyclic redundancy check (CRC) attachment module 102. Channel coding such as convolutional coding and turbo coding, is then carried out by a coding module 104. The encoded data is then interleaved in an interleaving module 106 in order to reduce burst errors in the data. Symbol mapping is carried out by a mapping module 108 on the interleaved data. A pilot signal is inserted by an insertion module 110 to aid in channel estimation at the receiver.

Signals out of the pilot insertion block 110 are applied to the serial to parallel conversion block 112 to become parallel signals. The said parallel signals are then OFDM modulated, i.e. inverse fast Fourier transformed (IFFT), by IFFT 114 and parallel to serial transformed by Parallel to serial block 116 to become OFDM symbols. The OFDM modulation can be expressed as follows:

where K is the total number of sub-carriers, s{k) is the complex symbol to be transmitted by the k-th sub-carrier, S(n) is the n-th time domain sample of the OFDM symbol, and j = V-T denotes the unit of the imaginary part of a complex symbol.

After OFDM modulation is performed, a Cyclic-prefix signal is added to the beginning of each OFDM symbol. The said Cyclic-prefix signal is a replica of the last L samples of S(n) as shown in Eq (1). Signals out of 118 are then digital-to-analogue converted (DAC) by 120 and transmitted by antenna 122. Note that the pre-antenna transmission processing after 122, such as amplifying the analogue signal and modulating the analogue signal to the carrier frequency is integrated in block 122.

Figure 2 shows a general diagram of the OFDM-based MIMO receiver, comprising receive antennas 202, spatial/space-time processor 316, and demodulation / decoding module 318. The detailed signal processing involved is given as follows: transmitted signals are received by receiving antennas 202, 202' at the receiver 200. The received signals go through another series of processes, which generally are the reverse of the processes carried out at the transmitter 100. The processes for one branch will be described. The initial reverse processes are analog to digital conversion by a module 204, removal of cyclic prefix by a module 206 and fast Fourier transformation (FFT) by a module 210. Since the received signals are comprised of overlapping signals from the multiple transmit antennas, it is necessary to separate the signals into their respective streams. A MIMO decoder 214, which makes use of zero forcing (ZF) or minimum mean square error (MMSE) techniques, is employed to perform this function, in the example embodiment. This is then followed by the processes of demapping performed by a module 216, deinterleaving performed by a module 218 and decoding performed by a module 220.

After decoding a cyclic redundancy check (CRC) is performed by a module 222 on each packet to validate the data. If the packet checked is error-free, an acknowledgement (ACK) is sent to the transmitter 100, and the transmitter 100 will not retransmit the packet. Otherwise, a no-acknowledgement (NACK) will be sent to the transmitter 100 to request retransmission. The feedback signal (ACK or NACK) can be transmitted through the reverse link, i.e. the path from the receive antenna 202 to the transmit antenna 122. A single data output digital signal 224 is thus recovered.

Figure 3 shows the transmit power allocation architecture of a MIMO OFDM communication system in an example embodiment. Techniques are provided to:(1) measure the channel link conditions via channel variance calculation, (2) feedback channel link information to transmitter through the previously described reverse link. (3) derive sub-carrier group size based on channel feedback, (4) calculate the number of sub-carrier groups, and (5) allocate transmit power in two proposed approaches, i.e. spatial-frequency and frequency-spatial methods.

In Figure 3, a transmitter block 302 and a receiver block 314 comprise of several modules devised for transmit power allocation using channel link condition feedback in an example embodiment. At the receiver block 314, multiple signals received via multiple receiving antennas 322 are processed via spatial / space-time MIMO processor 316. The MIMO decoded data packets are passed to the next processing block, i.e. the demodulation / decoding module 318. In this module 318, multiple streams of data packets are further demodulated and decoded before output. The detailed signal processing in blocks 316 and 318 is as described above with reference to Figure 2.

In order to measure multiple different link conditions, a module 320 measures channel link quality of the individual receiving antennas 322 via a channel variance calculation over K sub-carriers. The detailed definition of channel variance is given in Equation (3) for the example embodiment. The largest channel variance among multiple channels variances is selected as feedback to transmitter 302 via feedback channel 322.

As illustrated in Figure 3, the feedback value is sent to a module 304 where the setting of sub-carrier group size is performed. In the module 304, group size (number of sub-carriers in a sub-carrier group) is derived using a threshold method (described below in detail) in the example embodiment. Different group sizes are used according to different variance ranges, i.e. a smaller group size is used for a higher variance and vice versa. The group size thus derived is common to all antenna branches. Subsequent power allocation will be performed in terms of sub-carrier groups instead of individual sub-carriers for each antenna branch, in the example embodiment.

In another example embodiment, individual channel variances are measured by the module 320 and are feedback to transmitter 302 for individual group size setting instead of choosing the largest channel variance. Each spatial antenna branch will have an individual group size for sub-carrier grouping. The advantage of this procedure is greater accuracy in terms of link adaptation to channel variation for each channel. However, a more complex computation will be involved. Subsequent power allocation will again be performed in terms of sub-carrier groups instead of individual sub-carriers for each antenna branch in such an embodiment.

The number of sub-carrier groups is derived in a calculation module 306. It is derived by dividing the total number of sub-carriers per antenna branch by the group size input obtained from the module 304. This sub-carrier group number parameter is then input to a processing unit 308 for transmit power allocation determination. The steps of deriving the group size and number of sub-carrier groups are further described with reference to Figure 4.

Figure 4 shows the procedure, performed by the calculation module 320, of calculating the sub-carrier group parameters for use in transmit power allocation in an example embodiment. Step 402 calculates the mean channel gain h ' for antenna /over all K sub-carriers by Equation (2) as below:

where h_k' denotes the channel gain of the k -th sub-carrier over antenna / .

Loop step 404 ensures that all sub-carriers of each antenna branch / are taken into account in the calculation in step 402. In step 406, the channel variance Vaη associated with antenna channel / is calculated by Equation (3) as follows:

Where the superscript * denotes the complex conjugate.

Loop step 408 ensures that the calculation of channel variance is repeated for all the antenna branches before proceeding to step 410. In step 410, the largest channel variance of all the antenna channel variance is selected and feedback to transmitter in step 412.

In step 414, the sub-carrier group size G_r is derived by the module 304 using a threshold-based method. The assumption made here is that under a high variance case, a smaller group size should be obtained. The number of sub-carrier groups GR₁₁₀ is calculated using Equation (4) in step 416:

GR₁₁₀ = ^- Eq (4)

In step 418, the calculated G₁- and GR_no parameters are then input to the transmit power allocation unit 308 for further use in either Spatial-Frequency or Frequency- Spatial power allocation procedures respectively by modules 309 and 310.

Referring again to Figure 3, two transmit power allocation methods are available, i.e. Spatial-Frequency transmit power allocation and Frequency-Spatial transmit power allocation in modules 309 and 310 respectively. Based on the sub-carrier group size and group number information input from the module 306, transmit power is allocated to input data packets according to the methods of the module 309 or the module 310. After power is allocated to data packets, the multiple antenna data streams are processed by the module 312 for spatial / space-time MIMO arrangement. In the example embodiment, module 312 includes al of the components described above with reference to Figure 1 , except the antennas. Subsequently, the MIMO signals are transmitted via the multiple transmit antennas 313.

Figure 5 and Figure 6 illustrate the two power allocation methods, Spatial- Frequency transmit power allocation and Frequency-Spatial transmit power allocation schemes respectively, which operate under different channel link conditions in the example embodiment.

In Figure 5, power allocation is performed in the spatial dimension first (step

502), followed by allocation across the frequency dimension (step 506) in an example embodiment. Under high frequency selective fading channel conditions, channel variance is high. As such, the sub-carrier group size is small and this leads to largerGi?,,₀. Power allocation across the spatial dimension has lower computational complexity due to the smaller number of transmit antenna, compared with the typical number of sub-carriers.

In step 504, the total system transmission power, p_t , is firstly distributed to each antenna branch and then to each sub-carrier group within one antenna branch In one embodiment, the power allocation among antenna branches follows the equal power p allocation law, where equal power— — is allocated to all the antenna branches; in

another embodiment, the power allocation among antenna branches follows the water- filling law, i.e. larger transmission power is allocated to the antenna with larger channel

K-I power defined byP_cΛ , = ∑\ h_k' |² . The power P_ant , allocated to antenna i can be

expressed as: [Please provide further details on the "water-filling" method]

Eq (5) where ξ is the commonly-called water-level with the integer

p f denoting the number of the active antennas, / = ^■ — —is the average SNR with W

denoting the bandwidth of the OFDM signal and N₀ the power spectral density of the noise, and the notation (x)⁺ = max(x,θ). The active antenna number ^r can be obtained in an iterative manner, i.e. first let r = N₁ and calculate Eq (5), if all P_{mt l} > 0 , the obtained results are the final results; if one P_anl , < 0 , then let r = N₁ - 1 , delete antenna / , and calculate P_ant , among the rest antennas; These operations continue until all the allocated power in the r antennas are greater or equal to zero. The corresponding r antennas are called the active antennas.

In the power allocation across the frequency dimension (step 506), P_{mt l} is further distributed to each sub-carrier group in step 508 according to the channel power

G_rg of the sub-carrier group defined by P_{ch l g} = ∑l fy_t l² - '^{n one} embodiment, the k=G_r (g-\)+\ transmit power allocated to antenna i can be equally allocated to all its sub-carrier groups; in another embodiment, the transmission power P_{mt l} can be allocated to each of its sub-carrier group by the water-filling method as described previously; in another embodiment, a threshold method can be applied. The detailed description of the said threshold method is given below:

1) A threshold for the sub-carrier group channel power is determined by the system requirement, e.g. if the system requires the bit error rate (BER) less than 1(T³ , then the required average SΝR for the said BER can be calculated. The corresponding average channel power of the sub-carrier group can be obtained as well. The said average channel power of the sub-carrier group can be used as the threshold.

2) If a sub-carrier group has channel power less than the given threshold, the transmit power for this group is set to zero, i.e. no data is transmitted on this sub-carrier group. The total transmit power P_αn, , for the given antenna is then equally distributed in the rest of the sub-carrier groups.

Through step 510, the transmit power allocated to each sub-carrier group is equally distributed to the sub-carriers within the sub-carrier group since the channel gains of the said sub-carriers within one sub-carrier group do not typically change much.

Loop step 512 checks that the procedure in step 510 is repeated for all the sub- carrier groups before proceeding to step 514. Loop step 514 checks that procedures in steps 506 to 512 are repeated for all transmit antennas.

Under low frequency selective fading channel conditions, channel variance is low. As such, the sub-carrier group size is large and this leads to a smallerGi?_no . In

Figure 6, power allocation is performed in the frequency dimension first (step 602), followed by allocation across the spatial dimension (step 606). Power allocation across frequency dimension is lower in complexity due to the smaller number of sub-carrier groups compared to the typical number of antennas.

In step 604, total system transmission power, P₁ is firstly distributed equally to each sub-carrier group for all antenna branches. The following power equation has to be satisfied in step 604:

GR_no

∑^p _g, = p_t ^Eq ⁽6⁾

where P₁ is the total system transmission power,

P_{g n} is the power of each sub-carrier group across all transmit antenna, GR₁₇₀ is the number of sub-carrier group per antenna branch.

Under the power allocation across the spatial dimension within each sub-carrier group (step 606), P_{g n} is further distributed to each sub-carrier group of each antenna branch i in step 608. In one embodiment, the power allocation in one sub-carrier group follows the previously described water-filling law; in another embodiment, the power allocation in one sub-carrier group follows the previously described threshold method. The following equation has to be satisfied in step 608:

N, Σ^p _g^_nl,, = ^p _g,_n , n = \...GR_no Eq (7)

1=1

where P_{g anl l} is the power of each sub-carrier group of each antenna branch,

P_{g n} is the power of each sub-carrier group across all transmit antennas, N₁ is the total number of transmit antennas, GR₁₁₀ is the number of sub-carrier group per antenna branch.

In step 610, P_{g ant l} is equally allocated to individual sub-carriers within the sub- carrier group associated with antenna i. The following equation has to be satisfied in step 610:

P_subc = ^P*f^LL Eq (8) r

where P_{g ant l} is the power of each sub-carrier group of each antenna branch,

P_subc is the power of each sub-carrier within one sub-carrier group, G_r is the sub-carrier group size in units of sub-carrier.

Loop step 612 checks that the procedure in step 610 is repeated for all the transmit antennas before proceeding to step 614. Loop step 614 checks that procedures in steps 608 to 612 are repeated for all sub-carrier groups.

The group size setting can be based on different channel variances in order to reflect individual channel link condition in another example embodiment, as mentioned above. As link condition varies for different channels, such an embodiment may be more adaptive and should yield better performance. However, the trade-off is computational complexity as the sub-carrier group size is different for each antenna branch.

Embodiments of the present invention can reduce the computational complexity for the power allocation in a MIMO OFDM system. For example, in a N₁ x N₁MMO system with K sub-carriers for each transmitted OFDM symbol, the conventional water- filling method needs up to N₁ x K units for power allocation; while for the example embodiment with equal power allocation in each sub-carrier group, the number of units

for power allocation is only N₁ x , which reduces the computational complexity yGR_no J by factor of GR_no . When the threshold method is used, an embodiment of the present invention can reduce the complexity further since in this case the calculation of the transmit power to the power allocation units is linear, i.e. to turn off the units with channel power less than the threshold and to allocate the power equally to the unites with channel power greater than or equal to the threshold.

Although in the previously described embodiments of the invention, equal power allocation is used for power distribution within a sub-carrier group of a given antenna, other power allocation schemes can still be used. For example, if P_{g ant} , (refer to Eq (8)) is the power allocated to the g-th sub-carrier group of antenna i, a proportional power allocation method can be applied for distributing the power to each sub-carriers, that is

for the l-th sub-carrier, the allocated power is P' = Kf P_{g ant l} , where P_{ch l g} ^'\s the ch,ι,g previously defined channel power for sub-carrier group g. In such an embodiment, the computational complexity of the power allocation can be reduced since the calculation of

P₁' is significantly easier than the water-filling method as shown in Eq (5).

The criteria to switch between spatial-frequency and frequency-spatial methods in the example embodiments depends on the number of sub-carrier groups and the number of transmit antennas. When the former is larger, frequency-spatial method will preferably be used; else if the later is larger, the spatial-frequency method will preferably be used. The low complexity methods in example embodiments aim to achieve better system performance than uniform power allocation methods and reduction in transmit power allocation complexity as compared to waterfilling methods.

Figure 7 shows a flow-chart illustrating a method for transmit power allocation in wireless communication systems having a plurality of antenna branches and a plurality of sub-carriers on each antenna branch, in accordance with an example embodiment. At step 700, a feedback signal based on measured link parameters of the respective antenna branches is provided. At step 702, sizes of sub-carrier groups are determined for the antenna branches based on the feedback signal. At step 704, numbers of the sub-carrier groups for the antenna branches are determined based on the respective sub-carrier group sizes. At step 706, the transmit power allocation is performed based on the determined sub-carrier groups for the respective antenna branches.

It will be appreciated by a person skilled in the art that numerous variations and/or modifications may be made to the present invention as shown in the specific embodiments without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects to be illustrative and not restrictive.

For example, instead of channel variance, other channel state information such as channel RMS delay, τ_rms, and coherence bandwidth may be used as feedback in transmit power allocation methods in different example embodiment.

Claims

1. A method for transmit power allocation in wireless communication systems having a plurality of antenna branches in both the transmitter and the receiver and a plurality of sub-carriers implemented for each antenna branch, the method comprising: providing a feedback signal based on measured link parameters of one or more of the respective antenna branches; determining sizes of sub-carrier groups for the antenna branches based on the feedback signal; determining numbers of the sub-carrier groups for the antenna branches based on the respective sub-carrier group sizes; and performing transmit power allocation based on the determined sub-carrier groups for the respective antenna branches.

2. The method as claimed in claim 1, wherein the link parameter of each antenna branch comprises one or more of a group consisting of variance, RMS delay, τ_rms, and coherence bandwidth over all sub-carriers of each antenna branch.

3. The method as claimed in claim 2, wherein the feedback signal is based on an extreme measured link parameter for all antenna branches, and one sub-carrier group size is determined for all antenna branches based on the extreme measured link parameter.

4. The method as claimed in claim 2, wherein the feedback signal is based on each measured link parameter for the respective antenna branches, and the group sizes for the respective antenna branches are individually determined.

5. The method as claimed in any one of the preceding claims, wherein the power allocation is first performed across all antenna branches, and then across all sub-carrier groups and all sub-carriers of each sub-carrier group for each antenna branch.

6. The method as claimed in claim 5, wherein the power allocation across all antenna branches is based on a measured link property of the antenna branches.

7. The method as claimed in claim 6, wherein the power allocation is based on the equal power allocation law.

8. The method as claimed in claim 6, wherein the power allocation is based on the water-filling power allocation law.

9. The method as claimed in any one of claims 5 to 8, wherein the power allocation across all sub-carrier groups for each antenna branch is based on mean channel gain values of the respective sub-carrier groups.

10. The method as claimed in claim 9, wherein the power allocation is based on is based on the equal power allocation law.

11. The method as claimed in claim 9, wherein the power allocation is based on the water-filling power allocation law.

12. The method as claimed in claim 9, wherein the power allocation comprises a threshold method.

13. The method as claimed in any one of claims 5 to 12, wherein power is equally allocated to all sub-carriers of each sub-carrier group for each antenna branch.

14. The method as claimed in any one of claims 1 to 4, wherein the power allocation is first performed across all sub-carrier groups for all antenna branches, and then across all antenna branches for each sub-carrier group and all sub-carriers for each sub-carrier group for each antenna branch.

15. The method as claimed in claim 14, wherein power is allocated equally to all sub-carrier groups for all antenna branches.

16. The method as claimed in claims 14 or 15, wherein the power allocation for each sub-carrier group across all antenna branches is performed based on threshold decisions or a water-filling method.

17. The method as claimed in claims 15 or 16, wherein power is equally allocated across all sub-carriers for each sub-carrier group for each antenna branch.

18. A wireless communication system comprising: a transmitter; a receiver; a plurality of antenna branches implemented in both the transmitter and receiver, each antenna branch arranged for a plurality of sub-carriers; a measurement unit for measuring link parameters of the respective antenna branches; a feedback unit providing a feedback signal based on the measured link parameters; a group size setting module determining sizes of sub-carrier groups for the antenna branches based on the feedback signal; a group number setting module for determining numbers of the sub-carrier groups for the antenna branches based on the respective sub-carrier group sizes; and a transmit power allocation processor performing transmit power allocation based on the sub-carrier groups for the respective antenna branches.

19. A transmitter for a wireless communication system, the transmitter comprising: a plurality of antenna branches implemented in the transmitter, each antenna branch arranged for a plurality of sub-carriers; a feedback unit for receiving a feedback signal based on measured link parameters of the respective antenna branches; a group size setting module determining sizes of sub-carrier groups for the antenna branches based on the feedback signal; a group number setting module for determining numbers of the sub-carrier groups for the antenna branches based on the respective sub-carrier group sizes; and a transmit power allocation processor performing transmit power allocation based on the sub-carrier groups for the respective antenna branches.

20.. A receiver for a wireless communication system, the receiver comprising: a plurality of antenna branches implemented in the receiver, each antenna branch arranged for a plurality of sub-carriers; a measurement unit for measuring link parameters of the respective antenna branches; a feedback unit for providing a feedback signal based on the measured link parameters for determining sizes of sub-carrier groups for the antenna branches based on the feedback signal and for determining numbers of the sub-carrier groups for the antenna branches based on the respective sub-carrier group sizes.