CN115664477B - Multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization - Google Patents
Multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization Download PDFInfo
- Publication number
- CN115664477B CN115664477B CN202211287214.2A CN202211287214A CN115664477B CN 115664477 B CN115664477 B CN 115664477B CN 202211287214 A CN202211287214 A CN 202211287214A CN 115664477 B CN115664477 B CN 115664477B
- Authority
- CN
- China
- Prior art keywords
- channel
- quantization
- max
- transformation matrix
- feedback
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000013139 quantization Methods 0.000 title claims abstract description 62
- 230000009466 transformation Effects 0.000 title claims abstract description 60
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000006835 compression Effects 0.000 title claims abstract description 31
- 238000007906 compression Methods 0.000 title claims abstract description 31
- 239000011159 matrix material Substances 0.000 claims abstract description 78
- 238000004891 communication Methods 0.000 claims abstract description 16
- 238000004088 simulation Methods 0.000 claims abstract description 6
- 239000013598 vector Substances 0.000 claims description 25
- 230000008569 process Effects 0.000 claims description 13
- 230000005540 biological transmission Effects 0.000 claims description 11
- 238000000354 decomposition reaction Methods 0.000 claims description 4
- 230000004044 response Effects 0.000 claims description 4
- 238000012512 characterization method Methods 0.000 claims description 3
- 238000011156 evaluation Methods 0.000 claims description 3
- 238000013461 design Methods 0.000 abstract description 3
- 230000006872 improvement Effects 0.000 description 9
- 230000000694 effects Effects 0.000 description 5
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000001413 cellular effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The invention discloses a multi-antenna channel compression and feedback method based on Karhunen-Loeve (KL) transformation and Lloyd-Max (LM) quantization, belonging to the field of wireless communication physical layer protocol design. Comprising the following steps: the mobile terminal constructs a KL transformation matrix and a corresponding inverse matrix according to multi-antenna historical channel data in the observation window, and compresses real-time channel information by using the KL transformation matrix; the mobile terminal carries out uniform quantization coding on the KL inverse transformation matrix, carries out LM quantization coding on the compressed channel information, constructs a coding result into a feedback bit stream according to a feedback protocol and sends the feedback bit stream to the base station terminal; the mobile terminal simulation base station decodes and reconstructs the KL transformation matrix and the compressed channel information, compares the reconstruction channel with the real-time channel, calculates the reconstruction performance, and judges whether the KL transformation matrix and the corresponding inverse matrix need to be updated according to the reconstruction performance; and the base station end receives the feedback bit stream sent by the mobile terminal, decodes the KL inverse transformation matrix and the compressed channel information from the feedback bit stream according to the feedback protocol, and reconstructs a downlink communication channel.
Description
Technical Field
The invention relates to the technical field of wireless communication physical layer protocol design, in particular to a multi-antenna channel state information compression and feedback method based on KL transformation and Lloyd-Max quantization.
Background
Real-time and accurate estimation and feedback of wireless channel state information plays an important role in realizing channel adaptive code modulation and multi-user interference coordination of a Frequency Division Duplex (FDD) multi-user multi-antenna wireless communication system. Along with the improvement of radio frequency of wireless communication and the increase of the number of antennas of a communication base station, the time-varying characteristic of channel state information is enhanced, the dimension is increased, and the conventional channel quantization and feedback method based on a fixed codebook faces the problems of difficult codebook design, large quantization error and lack of adaptability to specific communication scenes. In recent years, the neural network self-encoder is applied to compression and reconstruction of wireless channels, and the feedback delay of the channels is remarkably reduced; however, the training of the self-encoder depends on a large amount of channel data obtained by previous survey, and the channel reconstruction performance of the trained self-encoder cannot be guaranteed after the radio frequency environment changes, which seriously hinders the application of the self-encoder in a practical system.
Disclosure of Invention
1. Technical problem to be solved by the invention
Aiming at the problems of large feedback overhead and lack of scene self-adaption in the downlink channel estimation feedback of the existing multi-user multi-antenna frequency division duplex communication system, the invention provides a multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max (LM) quantization. In order to realize self-adaptive channel compression and feedback, the mobile terminal firstly simulates a base station to reconstruct a compressed and quantized channel before feeding back channel information and evaluates the channel reconstruction performance to determine whether the KL transformation matrix needs to be updated.
2. Technical proposal
In order to achieve the above purpose, the technical scheme provided by the invention is as follows: the multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization comprises the following steps:
S1: the mobile terminal constructs a KL transformation matrix psi and a corresponding inverse matrix phi according to multi-antenna historical channel information in an observation window, and compresses real-time channel information by using the KL transformation matrix psi;
s2: the mobile terminal carries out uniform quantization coding on the KL inverse transformation matrix phi, carries out LM quantization coding on the compressed channel information, constructs a coding result into a feedback bit stream according to a feedback protocol and sends the feedback bit stream to the base station terminal;
s3: the mobile terminal simulation base station decodes and reconstructs the KL transformation matrix and the compressed channel information according to the constructed feedback bit stream; comparing the reconstruction channel with the real-time channel, calculating the reconstruction performance, and judging whether the KL transformation matrix and the corresponding inverse matrix need to be updated according to the reconstruction performance;
s4: and the base station end receives the feedback bit stream sent by the mobile terminal, decodes the KL inverse transformation matrix and the compressed channel information from the feedback bit stream according to the feedback protocol, and reconstructs a downlink communication channel.
As a further improvement of the invention, the invention is a multi-antenna base station, the channel is a downlink multi-input single-output channel from the base station to the mobile terminal and is characterized by using a space domain response vector h, h is an N-dimensional complex vector, and N is the number of the base station antennas.
As a further improvement of the invention, the KL transformation matrix constructed in the step S1 is as followsWherein, S k and U k are diagonal matrices composed of k (0 < k < N) maximum singular values before channel covariance matrix singular value decomposition r=usu H and singular value matrices composed of corresponding singular vectors; correspondingly, the constructed KL inverse transformation matrix is Φ=u kSk; channel covariance matrix utilization/>Calculating; h -t is from the historical channel information set Ω= { h -T,h-T+1,…,h-1 }, T is the observation window length and t+.n.
As a further improvement of the present invention, in step S1, the compressed channel is h p =ψh.
As a further improvement of the present invention, in step S2, the uniform quantization encoding process of the KL inverse transform matrix Φ is: extracting the real part and the imaginary part coefficients of phi to form a real vector f, obtaining the maximum absolute value phi max of elements in the vector, and encoding phi max into B f -bit floating point numbers; uniformly discretizing the element in f within the range of [ -phi max,Φmax ] and expressing the element as B p bit unsigned integer, wherein the quantization interval is that
As a further improvement of the present invention, in step S2, the LM quantization coding process of the compressed channel information is: extracting real and imaginary coefficients of h p to form a real vector h pr, calculating standard deviation sigma of elements in h pr, and encoding sigma into B f bit floating point number; according to the assumption that the elements in h pr obey one-dimensional normal distribution, quantizing the elements in h pr by using an LM quantizing encoder and representing the quantized elements as B h -bit unsigned integers; wherein the LM quantizes the encoderIndividual decision threshold/>The individual quantized representation values are pre-calculated by the Lloyd-Max algorithm for a standard normal distribution and multiplied by σ.
As a further improvement of the present invention, in steps S2 and S4, the feedback protocol used is a bit stream transmission structure agreed in advance for the mobile terminal and the base station, and the uplink feedback channel used specifically is determined by the physical layer protocol of the communication system.
As a further improvement of the present invention, steps S3 and S4 employ a consistent decoding process, the decoding steps being: b f bit data corresponding to phi max is read and converted into floating point number according to the agreed bit stream transmission structureThen reading N multiplied by k multiplied by 2 multiplied by B p bit data corresponding to the KL inverse transformation matrix phi, and combining/>, using a uniform quantization decoderReconstructing the KL inverse transformation matrix to obtain phi r; b f bit data corresponding to sigma is read according to the agreed bit stream transmission structure and converted into floating point number/>Then reading k multiplied by 2 multiplied by B h bit data corresponding to the subspace channel representation h p, and reconstructing h p by using an LM quantization decoder to obtain h pr; wherein LM quantized decoder/>Individual decision threshold/>The quantized representation values are pre-calculated by Lloyd-Max algorithm for standard normal distribution and multiplied by/>Obtaining; the reconstructed channel h r＝Φrhpr is calculated.
As a further improvement of the present invention, in step S3, a correlation coefficient for reconstruction performance is usedTo evaluate, where h -i and h r,-i represent the previous i-th estimated channel and reconstructed channel, and L is the number of evaluations; when ρ is worse than the agreed performance threshold ρ th, re-compute ψ and Φ from the recent historical channel information set and update the quantized coding result of Φ.
3. Advantageous effects
Compared with the prior art, the technical scheme provided by the invention has the following beneficial effects:
(1) Compared with the traditional channel information feedback method based on the fixed codebook, the method is easier to implement, and different compression ratios can be realized by selecting different subspace sizes k and different quantization accuracies B f,Bp,Bh; meanwhile, the constructed KL transformation matrix depends on the latest historical channel information corresponding to different mobile terminals in different scenes, and can realize self-adaptive compression and feedback.
(2) Compared with a wireless channel compression and reconstruction method based on a neural network self-encoder, the method does not need a large amount of pre-surveyed channel data, and can achieve a compression ratio and reconstruction performance similar to or even better than those of the method.
Drawings
Fig. 1 is a schematic diagram of a downlink channel compression and feedback scenario of a cellular network frequency division duplex multi-user multi-antenna wireless communication system to which the present invention is applicable;
FIG. 2 is a flow chart of an implementation of the multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of the present invention;
FIG. 3 is a feedback bit stream structure diagram in an embodiment of the invention;
fig. 4 and fig. 5 are graphs of channel reconstruction performance results on measured channel data for the method according to the embodiment of the present invention.
Detailed Description
For a further understanding of the present invention, the present invention will be described in detail with reference to the drawings and examples.
The structures, proportions, sizes, etc. shown in the drawings are shown only in connection with the present disclosure, and are not intended to limit the scope of the invention, since any modification, variation in proportions, or adjustment of the size, etc. of the structures, proportions, etc. should be considered as falling within the spirit and scope of the invention, without affecting the effect or achievement of the objective. Also, the terms "upper", "lower", "left", "right", "middle", and the like are used herein for descriptive purposes only and are not intended to limit the scope of the invention for modification or adjustment of the relative relationships thereof, as they are also considered within the scope of the invention without substantial modification to the technical context.
Example 1
Referring to fig. 2, the specific steps of the multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization of the present invention are described, wherein the base station is a multi-antenna base station, the channel is a downlink multi-input single-output channel from the base station to the mobile terminal and is characterized by using a spatial response vector h, h is an N-dimensional complex vector, and N is the number of base station antennas.
Specifically, the method of the invention comprises the following specific steps:
S1: the mobile terminal constructs a KL transformation matrix psi and a corresponding inverse matrix phi according to multi-antenna historical channel information in an observation window, and compresses real-time channel information by using the KL transformation matrix psi;
In this step, the constructed KL transformation matrix is Wherein, S k and U k are diagonal matrices composed of k (0 < k < N) maximum singular values before channel covariance matrix singular value decomposition r=usu H and singular value matrices composed of corresponding singular vectors;
Correspondingly, the constructed KL inverse transformation matrix is Φ=u kSk;
channel covariance matrix utilization Calculating;
-t From the historical channel information set Ω= { h -T,h-T+1,…,h-1 }, T is the observation window length and t≡n.
Wherein the compressed channel is h p =ψh.
S2: the mobile terminal carries out uniform quantization coding on the KL inverse transformation matrix phi, carries out LM quantization coding on the compressed channel information, constructs a coding result into a feedback bit stream according to a feedback protocol and sends the feedback bit stream to the base station terminal;
In this step, the uniform quantization encoding process of the KL inverse transform matrix Φ is:
Extracting the real part and the imaginary part coefficients of phi to form a real vector f, obtaining the maximum absolute value phi max of elements in the vector, and encoding phi max into B f -bit floating point numbers;
Uniformly discretizing the element in f within the range of [ -phi max,Φmax ] and expressing the element as B p bit unsigned integer, wherein the quantization interval is that
The LM quantization coding process of the compressed channel information is as follows:
Extracting real and imaginary coefficients of h p to form a real vector h pr, calculating standard deviation sigma of elements in h pr, and encoding sigma into B f bit floating point number;
According to the assumption that the elements in h pr obey one-dimensional normal distribution, quantizing the elements in h pr by using an LM quantizing encoder and representing the quantized elements as B h -bit unsigned integers; wherein 2 B -1 decision thresholds and 2 B quantization representative values of the LM quantization encoder are pre-calculated by an Lloyd-Max algorithm for a standard normal distribution and multiplied by σ.
S3: the mobile terminal simulation base station decodes and reconstructs the KL transformation matrix and the compressed channel information according to the constructed feedback bit stream; comparing the reconstruction channel with the real-time channel, calculating the reconstruction performance, and judging whether the KL transformation matrix and the corresponding inverse matrix need to be updated according to the reconstruction performance;
in this step, a correlation coefficient for reconstruction performance is used To evaluate, where h -i and h r,-i represent the previous i-th estimated channel and reconstructed channel, and L is the number of evaluations; when ρ is worse than the agreed performance threshold ρ th, re-compute ψ and Φ from the recent historical channel information set and update the quantized coding result of Φ.
S4: and the base station end receives the feedback bit stream sent by the mobile terminal, decodes the KL inverse transformation matrix and the compressed channel information from the feedback bit stream according to the feedback protocol, and reconstructs a downlink communication channel.
In this step, the structure of the bit stream agreed in the feedback protocol is shown in fig. 3.
In this embodiment, steps S3 and S4 employ a consistent decoding process, the decoding steps being:
B f bit data corresponding to phi max is read and converted into floating point number according to the agreed bit stream transmission structure
Then reading Nxkx2XB p bit data corresponding to KL inverse transformation matrix phi, combining by using uniform quantization decoderReconstructing the KL inverse transformation matrix to obtain phi r;
B f bit data corresponding to sigma is read and converted into floating point number according to agreed bit stream transmission structure Then reading k×2×b bit data corresponding to the subspace channel characterization h p, and reconstructing h p by using an LM quantization decoder to obtain h pr;
Wherein 2 B -1 decision thresholds and 2 B quantization representative values of the LM quantization decoder are pre-calculated by the Lloyd-Max algorithm for a standard normal distribution and multiplied by Obtaining;
The reconstructed channel h r＝Φrhpr is calculated.
Example 2
In order to better verify and explain the technical effects achieved by the second embodiment of the invention, the experimental results are obtained by a scientific demonstration method in the embodiment so as to verify the true effects of the method.
First, an applicable feedback scenario is selected, referring to fig. 1, which shows a downlink channel compression and feedback scenario of an applicable cellular network Frequency Division Duplex (FDD) multi-user multi-antenna wireless communication system of the present invention. The number of the base station antennas is n=64, and the mobile terminal has a single antenna. According to the 5G NR physical layer protocol of 3GPP, a base station transmits a pilot signal CSI-RS, a terminal measures the CSI-RS and estimates channel state information, and the estimated channel state information needs to be reported to the base station. In this embodiment, the carrier frequency is 3.7GHz, the original channel state information h is the channel response on one OFDM subcarrier, n=64-dimensional complex vector, and the real part (I-path) and the imaginary part (Q-path) are represented by 32-bit precision floating point numbers.
Further, in combination with the embodiment flow of the above embodiment, in this embodiment, the subspace size is selected to be k=12, the history observation window size is set to be n=256, and the quantization accuracy B f＝32,Bp＝8,Bh =4. Meanwhile, the terminal is provided with a special buffer memory for storing the latest 256 channel estimation results.
The specific steps of the combination data are as follows:
s1: the mobile terminal reads the latest 256 channel estimation results from the cache to form a historical channel state information set omega= { h -256,h-255,…,h-1 }, constructs a KL transformation matrix psi and a corresponding inverse matrix phi according to omega, and compresses real-time channel information by using the KL transformation matrix psi;
Specifically, the constructed KL transformation matrix is Wherein S k and U k are channel covariance matrix singular value decomposition r=usu H, and the diagonal matrix composed of the first 12 largest singular values and the singular value matrix composed of the corresponding singular vectors; correspondingly, the constructed KL inverse transformation matrix is Φ=u kSk; channel covariance matrix utilization/>Calculating; for real-time channel h, the compressed channel is h p =ψh.
Step S2: the mobile terminal carries out uniform quantization coding on the KL inverse transformation matrix phi, carries out LM quantization coding on the compressed channel information, constructs a coding result into a feedback bit stream according to a feedback protocol and sends the feedback bit stream to the base station terminal;
Wherein, the uniform quantization coding process of the KL inverse transformation matrix phi is as follows:
(1) Extracting the real part and the imaginary part coefficients of phi to form a real vector f, obtaining the maximum absolute value phi max of elements in the vector, and encoding phi max into 32-bit floating point numbers;
(2) Uniformly discretizing the element in f within the range of [ -phi max,Φmax ] and expressing the element as 8-bit unsigned integer, wherein the quantization interval is
Further, the LM quantization coding process of the compressed channel information h p is:
(1) Extracting real and imaginary coefficients of h p to form a real vector h pr, calculating standard deviation sigma of elements in h pr, and encoding sigma into 32-bit floating point numbers;
(2) According to the assumption that the elements in h pr obey one-dimensional normal distribution, quantizing the elements in h pr by using an LM quantizing encoder and expressing the quantized elements as 4-bit unsigned integers; wherein 15 decision thresholds and 16 quantization representing values of the LM quantization encoder are pre-calculated by an Lloyd-Max algorithm aiming at standard normal distribution and are obtained by multiplying sigma;
Specifically, the decision threshold is ：σ*[-2.400803,-1.843532,-1.437139,-1.099286,-0.799550,-0.522404,-0.258222,0.000000,0.258222,0.522404,0.799550,1.099286,1.437139,1.843532,2.400803]; and the quantized representation value is ：σ*[-2.732589,-2.069017,-1.618046,-1.256231,-0.942340,-0.656759,0.388048,-0.128395,0.128395,0.388048,0.656759,0.942340,1.256231,1.618046,2.069017,2.732589].
Specifically, the bit stream transmission structure agreed in the feedback protocol is shown in fig. 3, wherein the length of a bit vector b kl corresponding to the KL inverse transformation matrix Φ after encoding is 1+B f+N×k×2×Bp =12321 bits, and the 1 st bit is the encoding start flag bit, and the value is 1; followed by encoded bitstreams of a plurality of compressed channels; a plurality of compressed channels share a standard deviation sigma, and the coding length of the compressed channels is 32 bits; the length of the bit vector B hp corresponding to each compression channel is k multiplied by 2 multiplied by B h =96 or 1+k multiplied by 2 multiplied by B h =97 bits, 1 bit in the 97-bit codes is a coding start flag bit, and the value is 0; in this embodiment, the uplink feedback channel used is a Physical Uplink Shared Channel (PUSCH) of the 5G NR protocol.
S3: the mobile terminal simulation base station decodes and reconstructs the KL transformation matrix and the compressed channel information according to the constructed feedback bit stream; comparing the reconstruction channel with the real-time channel, calculating the reconstruction performance, and judging whether the KL transformation matrix and the corresponding inverse matrix need to be updated according to the reconstruction performance;
s4: and the base station end receives the feedback bit stream sent by the mobile terminal, decodes the KL inverse transformation matrix and the compressed channel information from the feedback bit stream according to the feedback protocol, and reconstructs a downlink communication channel.
Specifically, steps S3 and S4 employ a consistent decoding process, where the decoding steps are:
(1) According to the agreed bit stream transmission structure, reading 32-bit data corresponding to phi max according to the flag bit information and converting the 32-bit data into floating point numbers Then 64×12×2×8=12288 bit data corresponding to KL inverse transform matrix Φ is read, and combined/>, using a uniform quantization decoderReconstructing the KL inverse transformation matrix to obtain phi r;
(2) Reading 32-bit data corresponding to sigma according to the agreed bit stream transmission structure and converting the 32-bit data into floating point number according to the flag bit information Then, 12×2×4=128-bit data corresponding to the subspace channel characterization h p is read, and h p is reconstructed by using an LM quantization decoder to obtain h pr;
wherein 15 decision thresholds and 16 quantized representation values of the LM quantized decoder are pre-computed by the Lloyd-Max algorithm for a standard normal distribution and multiplied by Obtaining;
Specifically, the decision threshold is:
The quantization representation values are:
(3) The reconstructed channel h r＝Φrhpr is calculated.
Specifically, the reconstruction performance used in step S3 isWhere h -i and h r,-i represent the first i estimated channels and reconstructed channels, l=10 is the estimated number; when ρ is worse than the agreed performance threshold ρ th =0.95, re-compute ψ and Φ from the latest historical channel information set and update the quantized coding result of Φ.
In this embodiment, the overall compression rate of the channel information is:
Where M is the number of channels compressed by a particular KL transform matrix. When the channel subspace changes faster (for example, when the terminal moving speed is faster), the KL transformation matrix is updated faster, and the corresponding M is smaller; when the channel subspace is unchanged (for example, the terminal position is fixed), the KL transformation matrix is not updated, M tends to infinity at the moment, and the overall compression rate kappa tends to be 42.23.
In this embodiment, the effect of the present invention is verified by combining the simulation of the measured channel data, and referring to fig. 4, the channel reconfiguration performance is a time-varying curve, the average channel reconfiguration performance is 0.99281, and the overall compression ratio is 40.96 in the case that the mobile terminal position is unchanged. Fig. 5 is a graph of the channel reconstruction performance over time for a mobile terminal at a mobile speed of 30km/h, with an average channel reconstruction performance of 0.98282 and an overall compression of 40.12. The method can accurately reconstruct the channel information under the condition of higher compression rate, and has adaptability to different scenes.
The invention and its embodiments have been described above by way of illustration and not limitation, and the invention is illustrated in the accompanying drawings and described in the drawings in which the actual structure is not limited thereto. Therefore, if one of ordinary skill in the art is informed by this disclosure, the structural mode and the embodiments similar to the technical scheme are not creatively designed without departing from the gist of the present invention.
Claims (9)
1. The multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization is characterized by comprising the following steps: the method comprises the following steps:
S1: the mobile terminal constructs a KL transformation matrix psi and a corresponding inverse matrix phi according to multi-antenna historical channel information in an observation window, and compresses real-time channel information by using the KL transformation matrix psi;
s2: the mobile terminal carries out uniform quantization coding on the KL inverse transformation matrix phi, carries out LM quantization coding on the compressed channel information, constructs a coding result into a feedback bit stream according to a feedback protocol and sends the feedback bit stream to the base station terminal;
s3: the mobile terminal simulation base station decodes and reconstructs the KL transformation matrix and the compressed channel information according to the constructed feedback bit stream; comparing the reconstruction channel with the real-time channel, calculating the reconstruction performance, and judging whether the KL transformation matrix and the corresponding inverse matrix need to be updated according to the reconstruction performance;
s4: and the base station end receives the feedback bit stream sent by the mobile terminal, decodes the KL inverse transformation matrix and the compressed channel information from the feedback bit stream according to the feedback protocol, and reconstructs a downlink communication channel.
2. The multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of claim 1, wherein: the base station is a multi-antenna base station, the channel is a downlink multi-input single-output channel from the base station to the mobile terminal and is characterized by a space domain response vector h, h is an N-dimensional complex vector, and N is the number of base station antennas.
3. The multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of claim 1, wherein: in the step S1 of the process,
The constructed KL transformation matrix isWherein S k and U k are a diagonal matrix composed of k (0 < k < n) maximum singular values before channel covariance matrix singular value decomposition r=usu H and a singular value matrix composed of corresponding singular vectors;
Correspondingly, the constructed KL inverse transformation matrix is Φ=u kSk;
channel covariance matrix utilization Calculating;
-t From the historical channel information set Ω= { h -T,h-T+1,…,h-1 }, T is the observation window length and t≡n.
4. The multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of claim 1, wherein: in step S1, the compressed channel is h p =ψh.
5. The multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of claim 1, wherein: in the step S2, the uniform quantization encoding process of the KL inverse transformation matrix Φ is as follows:
Extracting the real part and the imaginary part coefficients of phi to form a real vector f, obtaining the maximum absolute value phi max of elements in the vector, and encoding phi max into B f -bit floating point numbers;
Uniformly discretizing the element in f within the range of [ -phi max,Φmax ] and expressing the element as B p bit unsigned integer, wherein the quantization interval is that
6. The multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of claim 1, wherein: in step S2, the LM quantization coding process of the compressed channel information is:
Extracting real and imaginary coefficients of h p to form a real vector h pr, calculating standard deviation sigma of elements in h pr, and encoding sigma into B f bit floating point number;
According to the assumption that the elements in h pr obey one-dimensional normal distribution, quantizing the elements in h pr by using an LM quantizing encoder and representing the quantized elements as B h -bit unsigned integers;
Wherein 2 B -1 decision thresholds and 2 B quantization representative values of the LM quantization encoder are pre-calculated by an Lloyd-Max algorithm for a standard normal distribution and multiplied by σ.
7. The multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of claim 1, wherein: in steps S2 and S4, the feedback protocol used is a bit stream transmission structure agreed in advance for the mobile terminal and the base station, and the specifically used uplink feedback channel is determined by the physical layer protocol of the communication system.
8. The multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of claim 1, wherein: the steps S3 and S4 adopt a consistent decoding process, and the decoding steps are as follows:
B f bit data corresponding to phi max is read and converted into floating point number according to the agreed bit stream transmission structure
Then reading Nxkx2XB p bit data corresponding to KL inverse transformation matrix phi, combining by using uniform quantization decoderReconstructing the KL inverse transformation matrix to obtain phi r;
B f bit data corresponding to sigma is read and converted into floating point number according to agreed bit stream transmission structure
Then reading k×2×b bit data corresponding to the subspace channel characterization p, and reconstructing h p by using an LM quantization decoder to obtain h pr;
Wherein 2 B -1 decision thresholds and 2 B quantization representative values of the LM quantization decoder are pre-calculated by the Lloyd-Max algorithm for a standard normal distribution and multiplied by Obtaining;
The reconstructed channel h r＝Φrhpr is calculated.
9. The multi-antenna channel compression feedback method based on KL transform and Lloyd-Max quantization of claim 1, wherein: in step S3, the correlation coefficient for reconstruction performance usedTo evaluate, where h -i and h r,-i represent the previous i-th estimated channel and reconstructed channel, and L is the number of evaluations;
When ρ is worse than the agreed performance threshold ρ th, re-compute ψ and Φ from the recent historical channel information set and update the quantized coding result of Φ.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211287214.2A CN115664477B (en) | 2022-10-20 | 2022-10-20 | Multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211287214.2A CN115664477B (en) | 2022-10-20 | 2022-10-20 | Multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115664477A CN115664477A (en) | 2023-01-31 |
| CN115664477B true CN115664477B (en) | 2024-04-26 |
Family
ID=84989642
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211287214.2A Active CN115664477B (en) | 2022-10-20 | 2022-10-20 | Multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115664477B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009094520A1 (en) * | 2008-01-25 | 2009-07-30 | Smith International, Inc. | Data compression method for use in downhole applications |
| CN110311718A (en) * | 2019-07-05 | 2019-10-08 | 东南大学 | Quantization and inverse quantization method in a kind of extensive mimo channel status information feedback |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200162142A1 (en) * | 2018-11-15 | 2020-05-21 | Samsung Electronics Co., Ltd. | Method and apparatus to enable csi reporting in wireless communication systems |
-
2022
- 2022-10-20 CN CN202211287214.2A patent/CN115664477B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2009094520A1 (en) * | 2008-01-25 | 2009-07-30 | Smith International, Inc. | Data compression method for use in downhole applications |
| CN110311718A (en) * | 2019-07-05 | 2019-10-08 | 东南大学 | Quantization and inverse quantization method in a kind of extensive mimo channel status information feedback |
Non-Patent Citations (2)
| Title |
|---|
| On sparsity of eigenportfolios to reduce transaction cost;Anqi Xiong;Ali N. Akansu;《Journal of Capital Markets Studies》;20190708;第3卷(第1期);82-90 * |
| 基于DCT的SAR原始数据压缩算法分析;曾尚春;朱兆达;;电讯技术;20090328(03);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115664477A (en) | 2023-01-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114128162B (en) | Method and apparatus for enhanced CSI reporting | |
| US9124328B2 (en) | System and method for channel information feedback in a wireless communications system | |
| US7330701B2 (en) | Low complexity beamformers for multiple transmit and receive antennas | |
| US8023457B2 (en) | Feedback reduction for MIMO precoded system by exploiting channel correlation | |
| EP3682568B1 (en) | Device and method for compressing and/or decompressing channel state information | |
| US8630311B2 (en) | System and method for reporting quantized feedback information for adaptive codebooks | |
| US8432988B2 (en) | System and method for quantization of channel state information | |
| US11159218B2 (en) | System and method for feeding back channel information | |
| US20110243207A1 (en) | System and Method for Adapting Codebooks | |
| KR20120048025A (en) | Method of identifying a precoding matrix corresponding to a wireless network channel and method of approximating a capacity of a wireless network channel in a wireless network | |
| CN117639860A (en) | Method and communication device for indicating and determining precoding vector | |
| CN113098804A (en) | Channel state information feedback method based on deep learning and entropy coding | |
| KR101481982B1 (en) | Feedback method and system of correlation matrix for antenna array | |
| CN102025450B (en) | Method for feeding back channel information coding and mobile terminal | |
| CN115664477B (en) | Multi-antenna channel compression feedback method based on KL transformation and Lloyd-Max quantization | |
| US20110058506A1 (en) | Device and method for transmitting channel information in wireless communication system | |
| WO2022233230A1 (en) | Method and system for channel state information feedback using sub-codebook based trellis coded quantization | |
| CN112236961B (en) | Channel state information feedback | |
| Yeung et al. | Enhanced trellis based vector quantization for coordinated beamforming | |
| CN112205049B (en) | Channel state information feedback | |
| KR20100028859A (en) | Apparatus and method for adaptive coloring codebook in a multiple input multiple output wireless communication system | |
| CN115133967B (en) | Data processing method, device, electronic equipment and storage medium | |
| CN111010218A (en) | Method for indicating and determining precoding vector and communication device | |
| Ndiaye et al. | Feedback Overhead in the FDD Downlink Massive MIMO System | |
| CN116346279A (en) | Information processing method, device, terminal and network equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |