CN113489510A - Large-scale MU-MIMO system DBP architecture signal detection method based on user grouping - Google Patents

Large-scale MU-MIMO system DBP architecture signal detection method based on user grouping Download PDF

Info

Publication number
CN113489510A
CN113489510A CN202110930486.9A CN202110930486A CN113489510A CN 113489510 A CN113489510 A CN 113489510A CN 202110930486 A CN202110930486 A CN 202110930486A CN 113489510 A CN113489510 A CN 113489510A
Authority
CN
China
Prior art keywords
dbp
user group
user
scale
antenna
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.)
Granted
Application number
CN202110930486.9A
Other languages
Chinese (zh)
Other versions
CN113489510B (en
Inventor
戴晓明
李华
罗裕全
闫甜甜
马心阳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Science and Technology Beijing USTB
Original Assignee
University of Science and Technology Beijing USTB
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by University of Science and Technology Beijing USTB filed Critical University of Science and Technology Beijing USTB
Priority to CN202110930486.9A priority Critical patent/CN113489510B/en
Publication of CN113489510A publication Critical patent/CN113489510A/en
Application granted granted Critical
Publication of CN113489510B publication Critical patent/CN113489510B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H04B7/0452—Multi-user MIMO systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/0202—Channel estimation
    • H04L25/0204—Channel estimation of multiple channels
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/0202—Channel estimation
    • H04L25/024—Channel estimation channel estimation algorithms
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/0202—Channel estimation
    • H04L25/024—Channel estimation channel estimation algorithms
    • H04L25/0256—Channel estimation using minimum mean square error criteria
    • 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

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Power Engineering (AREA)
  • Radio Transmission System (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

The invention provides a large-scale MU-MIMO system DBP architecture signal detection method based on user grouping, and belongs to the technical field of communication. The method comprises the following steps: aiming at a large-scale MU-MIMO system DBP framework, a base station antenna is divided into a plurality of antenna clusters, users are divided into a plurality of user groups, each antenna cluster independently and parallelly detects signals sent by each user group, and detected local information is sent to a central processing unit, wherein the detected local information comprises: characteristic information of a transmission signal of each user group; the central processing unit combines the local information of the antenna cluster based on the message mixing strategy and estimates the sending signals of each user group. By adopting the invention, the problem of serious performance reduction caused by the increase of the number of antenna clusters and the limitation of global information sharing in the traditional distributed signal detection algorithm based on the DBP framework can be obviously improved.

Description

Large-scale MU-MIMO system DBP architecture signal detection method based on user grouping
Technical Field
The invention relates to the technical field of communication, in particular to a DBP (direct bus protocol) architecture signal detection method of a large-scale MU-MIMO (multi-user multiple input multiple output) system based on user grouping.
Background
In a large-scale Multi-user Multiple-Input Multiple-Output (MU-MIMO) system, hundreds of antennas are assembled on a base station, and tens of users in the same frequency band are served simultaneously, so that the frequency spectrum efficiency and the energy efficiency are improved by one order of magnitude. The existing large-scale MU-MIMO system usually adopts a centralized baseband signal processing mode, and massive data needs to be transmitted from a radio frequency unit of a base station to a baseband processing unit. For example, in a 128-antenna massive MIMO system, employing a 10-bit analog-to-digital converter and operating at a 40MHz bandwidth, the raw baseband data rate exceeds 200 Gbps. Such high data rates exceed the bandwidth of existing high-speed interconnect standards (e.g., common public wireless interfaces) and reach the bandwidth and power consumption limits of existing chip input/output interfaces. Furthermore, existing detection algorithms are typically centralized at the base station, resulting in extremely high data transmission rates and computational complexity.
In the prior art, a Decentralized Baseband Processing (DBP) architecture (as shown in fig. 1) is proposed to alleviate bandwidth and computation bottleneck of centralized Baseband Processing, and a base station antenna is divided into a plurality of antenna clusters, where each antenna cluster includes a local rf circuit, a channel estimator, a signal detector and an associated computation unit, and performs signal Processing independently and in parallel, and transmits local information of each cluster to a central Processing unit. The centralized processing unit combines the messages according to a certain message mixing rule and further obtains an estimated symbol. Existing DBP-based signal detection algorithms, such as minimum mean-square error (MMSE), zero-forcing (ZF), and approximate messaging (LAMA) detection algorithms, typically use two message mixing approaches in the central processing unit: one is a hybrid matched filter output; one is to take the error variance after equalization as the weight to carry out weighted summation to the estimated symbol of each antenna cluster.
The traditional de-centralization detection algorithm based on the DBP is high in complexity, the performance loss of the detection algorithm is serious along with the increase of the number of antenna clusters, the performance loss problem caused by global shared information limitation due to the antenna clusters is not fully considered although the bottleneck problems of ultrahigh data transmission and chip bandwidth are solved by adopting the DBP framework, and therefore a low-complexity and high-efficiency distributed signal detection algorithm needs to be designed.
Disclosure of Invention
The embodiment of the invention provides a DBP framework signal detection method of a large-scale MU-MIMO system based on user grouping, which can obviously improve the problem of serious performance reduction of the traditional DBP framework-based distributed signal detection algorithm caused by the increase of the number of clusters and the limitation of global information sharing. The technical scheme is as follows:
the embodiment of the invention provides a large-scale MU-MIMO system DBP architecture signal detection method based on user grouping, which comprises the following steps:
aiming at a large-scale MU-MIMO system DBP framework, a base station antenna is divided into a plurality of antenna clusters, users are divided into a plurality of user groups, each antenna cluster independently and parallelly detects signals sent by each user group, and detected local information is sent to a central processing unit, wherein the detected local information comprises: characteristic information of a transmission signal of each user group;
the central processing unit combines the local information of the antenna cluster based on the message mixing strategy and estimates the sending signals of each user group.
Further, for the DBP architecture of the large-scale MU-MIMO system, dividing the base station antenna into a plurality of antenna clusters, dividing the user into a plurality of user groups, and independently and parallelly detecting the signals sent by each user group by each antenna cluster includes:
aiming at a DBP (direct bus protocol) framework of a large-scale MU-MIMO (multiple user-multiple input multiple output) system, large-scale antennas configured by a base station are divided into C antenna clusters, and each antenna cluster comprises NcThe antenna, the base station received signal vector y and the noise vector n are respectively divided into
Figure BDA0003211141010000021
And
Figure BDA0003211141010000022
channel matrix H baseIs divided into rows
Figure BDA0003211141010000023
Wherein, ycFor the received signal of the c-th cluster, ncNoise representing cluster c, HcRepresenting a channel matrix of the c-th cluster, and a superscript T representing a matrix transposition;
dividing users into G user groups;
dividing the channel matrix of each antenna cluster into H by columnsc＝[Hc,1,…,Hc,g,…,Hc,G]Then y iscExpressed as:
Figure BDA0003211141010000024
wherein Hc,gChannel matrix, x, representing the g user group in the c clustergA signal transmitted for a g-th user group;
and each antenna cluster carries out independent and parallel signal detection, wherein the signal detection algorithm comprises the following steps: a linear detection algorithm or a non-linear messaging algorithm.
Further, the nonlinear message passing algorithm is a message passing algorithm based on expected value propagation, and the message update rule of the message passing algorithm based on expected value propagation is as follows:
Figure BDA0003211141010000031
Figure BDA0003211141010000032
wherein the content of the first and second substances,
Figure BDA0003211141010000033
and
Figure BDA0003211141010000034
respectively representing VNx at the time of the t-th iteration within the c-th clustergTo FN fc,nAnd messages passed in the reverse direction, VN representing variable nodes, FN representing functional nodes,
Figure BDA0003211141010000035
in order to be the a-priori information,
Figure BDA0003211141010000036
represents each group KgJoint constellation symbol vector set, K, for individual usersgFor each group of users, x \ xgDenotes x divided by xgOther transmitted signals, p (y)c,n∣xg) Representing a likelihood function.
Further, the characteristic information of the transmission signal includes: a covariance matrix and a mean vector;
assuming that the transmitted signal x is a continuous random gaussian vector and the message passing between VN and FN is approximated as a multivariate gaussian distribution, the characteristics of the passing message are characterized by covariance and mean, and the computation rule of the covariance matrix and mean vector is expressed as:
Figure BDA0003211141010000037
Figure BDA0003211141010000038
wherein the content of the first and second substances,
Figure BDA0003211141010000039
and
Figure BDA00032111410100000310
respectively representing the g user group VNx at the t iterationgTo FNfc,nThe upper scale H denotes the conjugate transpose, ec,nExpress identity matrix
Figure BDA00032111410100000322
N of (1)cIndicating the number of antennas in each antenna cluster,
Figure BDA00032111410100000311
and
Figure BDA00032111410100000312
the mean vector and covariance matrix of the approximate posterior distribution obtained for the moment matching respectively,
Figure BDA00032111410100000313
and
Figure BDA00032111410100000314
FNf at the t-1 th iterationc,nDelivery to the g-th user group VNxgMean and variance of.
Further, the mean vector of the approximate posterior distribution obtained by the moment matching
Figure BDA00032111410100000315
Sum covariance matrix
Figure BDA00032111410100000316
Respectively expressed as:
Figure BDA00032111410100000317
Figure BDA00032111410100000318
wherein the content of the first and second substances,
Figure BDA00032111410100000319
x is the t-th iteration time in the c-th clustergThe confidence of,
Figure BDA00032111410100000320
Is expressed at a confidence level of
Figure BDA00032111410100000321
Mean of time-variant.
Further, assuming that the transmitted signal x is a continuous random gaussian vector and the information between VN and FN is approximated as a multivariate gaussian distribution, the mean and variance of the FN delivered to VN are expressed as:
Figure BDA0003211141010000041
Figure BDA0003211141010000042
wherein the content of the first and second substances,
Figure BDA0003211141010000043
and
Figure BDA0003211141010000044
respectively representing FNf times of the t-th iterationc,nDelivery to the g-th user group VNxgMean and variance of Hc,g′A channel matrix representing the g' th user group of the c-th cluster,
Figure BDA0003211141010000045
and
Figure BDA0003211141010000046
respectively representing the g' th user group VNx at the t-th iterationg′To FNfc,nThe covariance matrix and the mean vector of (a),
Figure BDA0003211141010000047
is the variance of gaussian noise.
Further, the central processing unit combines the local information of the antenna cluster based on the message mixing strategy, and estimates the transmission signal of each user group includes:
the central processing unit combines the local information of each antenna cluster based on a multidimensional Gaussian multiplication message mixing strategy;
calculating the posterior probability of the signals sent by each user group according to the combined information;
and estimating the signals transmitted by each user group according to the obtained posterior probability of each group of user transmission signals.
Further, the merged global information α (x)g) Is approximated as a covariance matrix of
Figure BDA0003211141010000048
Mean vector of
Figure BDA0003211141010000049
Gaussian distribution of
Figure BDA00032111410100000410
Expressed as:
Figure BDA00032111410100000411
wherein the content of the first and second substances,
Figure BDA00032111410100000412
Figure BDA00032111410100000413
respectively representing the mean vector and covariance matrix for each user group when the maximum number of iterations is reached.
Further, the posterior probability of each user group sending signal is:
Figure BDA00032111410100000414
where ξ is the vector composed of the constellation symbols sent by each user group,
Figure BDA00032111410100000418
a set of constellation diagrams is represented,
Figure BDA00032111410100000415
represents each group KgJoint constellation vector set, β (x), for individual usersgξ) represents xgGlobal confidence in ξ, p (x)gξ) represents xgA priori of (a) (x)gξ) represents xgGlobal information in ξ.
Further, the estimated signals transmitted by each user group are:
Figure BDA00032111410100000416
wherein the content of the first and second substances,
Figure BDA00032111410100000417
indicating the estimated transmission signal of the g-th user group.
The technical scheme provided by the embodiment of the invention has the beneficial effects that at least:
in the embodiment of the invention, aiming at a DBP (direct bus protocol) framework of a large-scale MU-MIMO (multi-user multiple input multiple output) system, firstly, a base station antenna is divided into a plurality of antenna clusters, users are divided into a plurality of user groups, and each antenna cluster independently and parallelly detects signals sent by each user group; then, sending the detected local information to a central processing unit, wherein the detected local information comprises: characteristic information of a transmission signal of each user group; and finally, the central processing unit merges the local information of each antenna cluster based on a message mixing strategy and estimates the sending signals of each user group. By adopting the method and the device, the problem of serious performance reduction caused by the DBP architecture-based detection performance of the traditional large-scale MU-MIMO system along with the increase of the number of clusters and the limitation of global information sharing can be obviously improved.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a block diagram of a DBP architecture receiving end system;
fig. 2 is a schematic flowchart of a method for detecting DBP architecture signals of a large-scale MU-MIMO system based on user grouping according to an embodiment of the present invention;
fig. 3 is a factor graph of an EP detection algorithm based on user grouping according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.
As shown in fig. 1, an embodiment of the present invention provides a method for detecting a DBP architecture signal of a large-scale MU-MIMO system based on user grouping, including:
s101, aiming at a large-scale MU-MIMO system DBP framework, dividing a base station antenna into a plurality of antenna clusters, dividing a user into a plurality of user groups, independently and parallelly detecting signals sent by each user group by each antenna cluster, and sending detected local information to a central processing unit, wherein the detected local information comprises: characteristic information of a transmission signal of each user group;
s102, the central processing unit combines the local information of the antenna cluster based on the message mixing strategy and estimates the sending signals of all user groups.
Aiming at the DBP architecture of the large-scale MU-MIMO system, firstly, dividing a base station antenna into a plurality of antenna clusters, dividing a user into a plurality of user groups, and independently and parallelly detecting signals sent by the user groups by each antenna cluster; then, sending the detected local information to a central processing unit, wherein the detected local information comprises: characteristic information of a transmission signal of each user group; and finally, the central processing unit merges the local information of each antenna cluster based on a message mixing strategy and estimates the sending signals of each user group. By adopting the method and the device, the problem of serious performance reduction caused by the DBP architecture-based detection performance of the traditional large-scale MU-MIMO system along with the increase of the number of antenna clusters and the limitation of global information sharing can be obviously improved.
Fig. 1 shows a DBP architecture receiving end system in this embodiment, KAnd transmitting information to the base station provided with the N antennas by the single-antenna user in the same time-frequency resource. The source bit sequence transmitted by user enters into interleaver after channel coding, every Q coded bits are modulated into a complex constellation point with energy normalization (the constellation set is
Figure BDA0003211141010000061
And is
Figure BDA0003211141010000062
). All users' transmitted signals (vector in nature)
Figure BDA0003211141010000063
Can be expressed as x ═ x1,x2,…,xK]TThen, the complex baseband input-output model of the large-scale MU-MIMO system can be expressed as:
y＝Hx+n
wherein the content of the first and second substances,
Figure BDA0003211141010000064
a received signal vector representing a dimension of the base station side as Nx 1;
Figure BDA0003211141010000065
representing a rayleigh flat fading channel matrix, wherein each element of the channel matrix obeys a complex gaussian distribution with a mean value of zero and a variance of 1;
Figure BDA0003211141010000066
representing additive white Gaussian noise vector, the elements of which are independent and identically distributed circularly symmetric complex Gaussian random variables, the mean value of which is 0, and the variance of which is sigma2。
In a specific implementation manner of the foregoing large-scale MU-MIMO system DBP architecture signal detection method based on user grouping, further, for the large-scale MU-MIMO system DBP architecture, dividing a base station antenna into a plurality of antenna clusters, dividing a user into a plurality of user groups, where each antenna cluster independently and parallelly detects a signal sent by each user group includes:
a1, for DBP architecture of large-scale MU-MIMO system, dividing large-scale antennas into C antenna clusters at base station end, assuming the number of antennas in each antenna cluster is the same and NcThe base station receiving signal vector y and the noise vector n are respectively divided into
Figure BDA0003211141010000067
And
Figure BDA0003211141010000068
channel matrix H is divided into based on rows
Figure BDA0003211141010000069
Wherein, ycFor the received signal of the c-th antenna cluster (c-th cluster for short), ncNoise representing cluster c, HcRepresenting a channel matrix of the c-th cluster, and a superscript T representing a matrix transposition;
in this embodiment, when the number of antennas is equal to or greater than 64, the antenna may be referred to as a large-scale antenna.
A2, dividing users into G user groups, wherein the number of each user group is Kg；
A3, dividing the channel matrix of each antenna cluster into H by columnsc＝[Hc,1,…,Hc,g,…,Hc,G]Then y iscExpressed as:
Figure BDA00032111410100000610
wherein the content of the first and second substances,
Figure BDA0003211141010000071
a channel matrix representing the g user group of the c-th cluster,
Figure BDA0003211141010000072
a transmission signal for the g-th user group;
a4, performing independent and parallel signal detection on each antenna cluster, where the signal detection algorithm may adopt a linear detection algorithm, such as a minimum mean square error algorithm, or may also adopt a non-linear algorithm, such as a Message Passing Algorithm (MPA), and specifically may be: message passing algorithms based on expected value propagation, other algorithms are also possible in practical applications.
It should be noted that different algorithms have different local information expression forms, such as covariance matrix, mean vector, and gray matrix, and the local information generated by different algorithms is different.
The present embodiment takes an MPA detection algorithm based on Expected Propagation (EP) as an example to describe the signal detection process of each antenna cluster.
In this embodiment, when the EP-based MPA detection scheme is used to perform signal detection, the obtained local information (i.e., the characteristic information of the transmission signal) includes: a covariance matrix and a mean vector; an EP-based MPA detection scheme may be represented by a factor graph shown in fig. 3, where N is 8, K is 4, C is 2, and G is 2, and the message update rule of EP-based MPA is as follows:
Figure BDA0003211141010000073
Figure BDA0003211141010000074
wherein the content of the first and second substances,
Figure BDA0003211141010000075
and
Figure BDA0003211141010000076
respectively representing Variable Node (VN) x at the t iteration in the c clustergTo Function Node (FN) fc,nAnd a message to be delivered in the reverse direction,
Figure BDA0003211141010000077
in order to be the a-priori information,
Figure BDA0003211141010000078
represents each group KgJoint constellation symbol vector set, K, for individual usersgFor each group of users, x \ xgDenotes x divided by xgOther transmitted signals, p (y)c,n∣xg) Representing a likelihood function, defining a maximum number of iterations as TmaxGenerally, take Tmax＝6。
According to the characteristics of the EP algorithm, assuming that the transmitted signal x is a continuous random gaussian vector and the message transmitted between VN and FN is approximated to a multivariate gaussian distribution, the characteristics of the transmitted message can be characterized by covariance and mean, and the computation rule of the covariance matrix and mean vector is expressed as:
Figure BDA0003211141010000079
Figure BDA00032111410100000710
wherein the content of the first and second substances,
Figure BDA0003211141010000081
and
Figure BDA0003211141010000082
respectively representing the g user group VNx at the t iterationgTo FNfc,nThe upper scale H denotes the conjugate transpose, ec,nExpress identity matrix
Figure BDA0003211141010000083
N of (1)cIndicating the number of antennas in each antenna cluster,
Figure BDA0003211141010000084
and
Figure BDA0003211141010000085
the mean vector and covariance matrix of the approximate posterior distribution obtained for the moment matching respectively,
Figure BDA0003211141010000086
and
Figure BDA0003211141010000087
FNf at the t-1 th iterationc,nDelivery to the g-th user group VNxgMean and variance of; wherein the mean vector of the approximate posterior distribution obtained by the moment matching
Figure BDA0003211141010000088
Sum covariance matrix
Figure BDA0003211141010000089
Respectively expressed as:
Figure BDA00032111410100000810
Figure BDA00032111410100000811
wherein the content of the first and second substances,
Figure BDA00032111410100000812
for the t-th iteration x in the c-th antenna clustergThe confidence of,
Figure BDA00032111410100000813
Is expressed at a confidence level of
Figure BDA00032111410100000814
Mean of time-variant.
The mean and variance of FN delivery to VN are expressed as:
Figure BDA00032111410100000815
Figure BDA00032111410100000816
wherein the content of the first and second substances,
Figure BDA00032111410100000817
and
Figure BDA00032111410100000818
respectively representing FNf times of the t-th iterationc,nDelivery to the g-th user group VNxgMean and variance of Hc,g′A channel matrix representing the g' th user group of the c-th cluster,
Figure BDA00032111410100000819
and
Figure BDA00032111410100000820
respectively representing the g' th user group VNx at the t-th iterationg′To FNfc,nThe covariance matrix and the mean vector of (a),
Figure BDA00032111410100000821
is the variance of gaussian noise.
When the detection is finished, namely: detecting each antenna cluster to reach the maximum iteration number TmaxThe covariance matrix of each user group detected by each antenna cluster is determined
Figure BDA00032111410100000822
Sum mean vector
Figure BDA00032111410100000823
And sending the data to a central processing unit.
In a specific implementation manner of the foregoing method for detecting a DBP architecture signal of a large-scale MU-MIMO system based on user grouping, further, the merging, by the central processing unit, local information of an antenna cluster and estimating a signal sent by each user group based on a multidimensional gaussian-multiplied message mixing strategy includes:
b1, the central processing unit combines the local information of each antenna cluster based on the multidimensional Gaussian multiplied message mixing strategy;
in this embodiment, inMerging Node (Sum Node, SN) and merging the global information alpha (x)g) Approximately obey a covariance matrix of
Figure BDA00032111410100000824
Mean vector of
Figure BDA00032111410100000825
Gaussian distribution of
Figure BDA00032111410100000826
Expressed as:
Figure BDA00032111410100000827
wherein the content of the first and second substances,
Figure BDA00032111410100000828
b2, calculating the posterior probability of the signals sent by each user group according to the combined information;
in this embodiment, the posterior probability of each user group sending a signal is:
Figure BDA0003211141010000091
where ξ is the constellation symbol vector sent by each user group,
Figure BDA0003211141010000092
a set of constellation diagrams is represented,
Figure BDA0003211141010000093
represents each group KgSet of joint constellation symbol vectors, β (x), for individual usersgξ) represents xgGlobal confidence in ξ, p (x)gξ) denotes the prior probability, α (x)gξ) represents xgGlobal information in ξ.
And B3, estimating the signals transmitted by each user group according to the obtained posterior probability of each user transmitted signal.
In this embodiment, the estimated signals sent by each user group are:
Figure BDA0003211141010000094
wherein the content of the first and second substances,
Figure BDA0003211141010000095
indicating the estimated transmission signal of the g-th user group.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (10)

1. A large-scale MU-MIMO system DBP architecture signal detection method based on user grouping is characterized by comprising the following steps:
aiming at a large-scale MU-MIMO system DBP framework, a base station antenna is divided into a plurality of antenna clusters, users are divided into a plurality of user groups, each antenna cluster independently and parallelly detects signals sent by each user group, and detected local information is sent to a central processing unit, wherein the detected local information comprises: characteristic information of a transmission signal of each user group;
the central processing unit combines the local information of the antenna cluster based on the message mixing strategy and estimates the sending signals of each user group.
2. The DBP architecture signal detection method for the large-scale MU-MIMO system based on the user grouping of claim 1, wherein the dividing of the base station antennas into a plurality of antenna clusters for the large-scale MU-MIMO system DBP architecture, the dividing of the users into a plurality of user groups, each antenna cluster independently detecting the signals transmitted by each user group in parallel comprises:
aiming at a DBP (direct bus protocol) framework of a large-scale MU-MIMO (multiple user-multiple input multiple output) system, a large-scale antenna configured by a base station is divided into C antenna clusters, and each antenna cluster isThe antenna cluster comprises NcThe antenna, the base station received signal vector y and the noise vector n are respectively divided into
Figure FDA0003211139000000011
And
Figure FDA0003211139000000012
channel matrix H is divided into based on rows
Figure FDA0003211139000000013
Wherein, ycFor the received signal of the c-th cluster, ncNoise representing cluster c, HcRepresenting a channel matrix of the c-th cluster, and a superscript T representing a matrix transposition;
dividing users into G user groups;
dividing the channel matrix of each antenna cluster into H by columnsc＝[Hc,1,…,Hc,g,…,Hc,G]Then y iscExpressed as:
Figure FDA0003211139000000014
wherein Hc,gChannel matrix, x, representing the g user group in the c clustergA signal transmitted for a g-th user group;
and each antenna cluster carries out independent and parallel signal detection, wherein the signal detection algorithm comprises the following steps: a linear detection algorithm or a non-linear messaging algorithm.
3. The DBP architecture signal detection method for a large-scale MU-MIMO system based on user grouping as claimed in claim 2, wherein said nonlinear message passing algorithm is an expectation-propagation based message passing algorithm, and the message update rule of the expectation-propagation based message passing algorithm is:
Figure FDA0003211139000000015
Figure FDA0003211139000000021
wherein the content of the first and second substances,
Figure FDA0003211139000000022
and
Figure FDA0003211139000000023
respectively representing VNx at the time of the t-th iteration within the c-th clustergTo FNfc,nAnd messages passed in the reverse direction, VN representing variable nodes, FN representing functional nodes,
Figure FDA0003211139000000024
in order to be the a-priori information,
Figure FDA0003211139000000025
represents each group KgJoint constellation symbol vector set, K, for individual usersgFor each group of users, x \ xgDenotes x divided by xgOther transmitted signals, p (y)c,n∣xg) Representing a likelihood function.
4. The DBP architecture signal detection method for a large-scale MU-MIMO system based on user grouping as claimed in claim 3, wherein the characteristic information of the transmitted signal comprises: a covariance matrix and a mean vector;
assuming that the transmitted signal x is a continuous random gaussian vector and the message passing between VN and FN is approximated as a multivariate gaussian distribution, the characteristics of the passing message are characterized by covariance and mean, and the computation rule of the covariance matrix and mean vector is expressed as:
Figure FDA0003211139000000026
Figure FDA0003211139000000027
wherein the content of the first and second substances,
Figure FDA0003211139000000028
and
Figure FDA0003211139000000029
respectively representing the g user group VNx at the t iterationgTo FNfc,nThe upper scale H denotes the conjugate transpose, ec,nExpress identity matrix
Figure FDA00032111390000000210
N of (1)cIndicating the number of antennas in each antenna cluster,
Figure FDA00032111390000000211
and
Figure FDA00032111390000000212
the mean vector and covariance matrix of the approximate posterior distribution obtained for the moment matching respectively,
Figure FDA00032111390000000213
and
Figure FDA00032111390000000214
FNf at the t-1 th iterationc,nDelivery to the g-th user group VNxgMean and variance of.
5. The method according to claim 4, wherein the vector of the mean value of the approximate posterior distribution obtained by the moment matching is used as a signal detection method for the DBP architecture of the large-scale MU-MIMO system based on the user grouping
Figure FDA00032111390000000215
Sum covariance matrix
Figure FDA00032111390000000216
Respectively expressed as:
Figure FDA00032111390000000217
Figure FDA00032111390000000218
wherein the content of the first and second substances,
Figure FDA00032111390000000219
x is the t-th iteration time in the c-th clustergThe confidence of,
Figure FDA00032111390000000220
Is expressed at a confidence level of
Figure FDA00032111390000000221
Mean of time-variant.
6. The DBP (wideband signal detection) method for a large-scale MU-MIMO system according to claim 4, wherein the transmitted signal x is assumed to be a continuous random Gaussian vector and the information between VN and FN is approximated as a multi-Gaussian distribution, and the mean and variance of FN transmitted to VN are expressed as:
Figure FDA0003211139000000031
Figure FDA0003211139000000032
wherein the content of the first and second substances,
Figure FDA0003211139000000033
and
Figure FDA0003211139000000034
respectively representing FNf times of the t-th iterationc,nDelivery to the g-th user group VNxgMean and variance of Hc,g′A channel matrix representing the g' th user group of the c-th cluster,
Figure FDA0003211139000000035
and
Figure FDA0003211139000000036
respectively representing the g' th user group VNx at the t-th iterationg′To FNfc,nThe covariance matrix and the mean vector of (a),
Figure FDA0003211139000000037
is the variance of gaussian noise.
7. The DBP architecture signal detection method for the large-scale MU-MIMO system based on user grouping as claimed in claim 1, wherein the said central processing unit combines the local information of the antenna cluster based on the message mixing strategy, and estimates the sending signal of each user group includes:
the central processing unit combines the local information of each antenna cluster based on a multidimensional Gaussian multiplication message mixing strategy;
calculating the posterior probability of the signals sent by each user group according to the combined information;
and estimating the signals transmitted by each user group according to the obtained posterior probability of each group of user transmission signals.
8. The DBP architecture signal detection method for large-scale MU-MIMO system based on user grouping as claimed in claim 7, wherein the merged global information α (x)g) Is approximated as a covariance matrix of
Figure FDA0003211139000000038
Mean vector of
Figure FDA0003211139000000039
Gaussian distribution of
Figure FDA00032111390000000310
Expressed as:
Figure FDA00032111390000000311
wherein the content of the first and second substances,
Figure FDA00032111390000000312
Figure FDA00032111390000000313
respectively representing the mean vector and covariance matrix for each user group when the maximum number of iterations is reached.
9. The DBP (direct bus) architecture signal detection method for the large-scale MU-MIMO system based on the user grouping as claimed in claim 7, wherein the posterior probability of the signals transmitted by each user group is:
Figure FDA00032111390000000314
where ξ is the vector composed of the constellation symbols sent by each user group,
Figure FDA00032111390000000315
a set of constellation diagrams is represented,
Figure FDA00032111390000000316
represents each group KgJoint constellation vector set, β (x), for individual usersgξ) represents xgGlobal in ξConfidence, p (x)gξ) represents xgA priori of (a) (x)gξ) represents xgGlobal information in ξ.
10. The DBP structure signal detection method of claim 8, wherein the estimated signals sent by each user group are:
Figure FDA0003211139000000041
wherein the content of the first and second substances,
Figure FDA0003211139000000042
indicating the estimated transmission signal of the g-th user group.
CN202110930486.9A 2021-08-13 2021-08-13 Large-scale MU-MIMO system DBP architecture signal detection method based on user grouping Active CN113489510B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110930486.9A CN113489510B (en) 2021-08-13 2021-08-13 Large-scale MU-MIMO system DBP architecture signal detection method based on user grouping

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110930486.9A CN113489510B (en) 2021-08-13 2021-08-13 Large-scale MU-MIMO system DBP architecture signal detection method based on user grouping

Publications (2)

Publication Number Publication Date
CN113489510A true CN113489510A (en) 2021-10-08
CN113489510B CN113489510B (en) 2022-09-02

Family

ID=77946483

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110930486.9A Active CN113489510B (en) 2021-08-13 2021-08-13 Large-scale MU-MIMO system DBP architecture signal detection method based on user grouping

Country Status (1)

Country Link
CN (1) CN113489510B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116015495A (en) * 2022-12-14 2023-04-25 东南大学 Distributed detection method based on Markov chain Monte Carlo

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107733487A (en) * 2017-10-11 2018-02-23 北京科技大学 A kind of signal detecting method and device of extensive multi-input multi-output system
CN107919895A (en) * 2017-11-16 2018-04-17 东南大学 A kind of Distributed Detection method of large-scale and multiple users mimo system
CN108512795A (en) * 2018-03-19 2018-09-07 东南大学 A kind of OFDM receiver baseband processing method and system based on low Precision A/D C
CN111510403A (en) * 2020-04-14 2020-08-07 重庆邮电大学 Large-scale MIMO system iterative signal detection method based on symmetry L Q
US20200274558A1 (en) * 2015-12-28 2020-08-27 China Academy Of Telecommunications Technology Method and apparatus for signal processing
CN112242861A (en) * 2020-09-01 2021-01-19 宁波大学 User activity and signal joint detection method of large-scale MIMO system
CN112787694A (en) * 2021-01-12 2021-05-11 重庆邮电大学 Low-complexity detection algorithm of MIMO-SCMA system based on expected propagation

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200274558A1 (en) * 2015-12-28 2020-08-27 China Academy Of Telecommunications Technology Method and apparatus for signal processing
CN107733487A (en) * 2017-10-11 2018-02-23 北京科技大学 A kind of signal detecting method and device of extensive multi-input multi-output system
CN107919895A (en) * 2017-11-16 2018-04-17 东南大学 A kind of Distributed Detection method of large-scale and multiple users mimo system
CN108512795A (en) * 2018-03-19 2018-09-07 东南大学 A kind of OFDM receiver baseband processing method and system based on low Precision A/D C
CN111510403A (en) * 2020-04-14 2020-08-07 重庆邮电大学 Large-scale MIMO system iterative signal detection method based on symmetry L Q
CN112242861A (en) * 2020-09-01 2021-01-19 宁波大学 User activity and signal joint detection method of large-scale MIMO system
CN112787694A (en) * 2021-01-12 2021-05-11 重庆邮电大学 Low-complexity detection algorithm of MIMO-SCMA system based on expected propagation

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
ZHENYU ZHANG: "Decentralized Signal Detection via Expectation Propagation Algorithm for Uplink Massive MIMO Systems", 《IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY》 *
张振宇: "面向6G超大规模MIMO***分布式基带处理信号检测算法", 《移动通信》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116015495A (en) * 2022-12-14 2023-04-25 东南大学 Distributed detection method based on Markov chain Monte Carlo

Also Published As

Publication number Publication date
CN113489510B (en) 2022-09-02

Similar Documents

Publication Publication Date Title
Jeon et al. Blind detection for MIMO systems with low-resolution ADCs using supervised learning
Liu et al. Optimized uplink transmission in multi-antenna C-RAN with spatial compression and forward
Nguyen et al. SVM-based channel estimation and data detection for one-bit massive MIMO systems
Li et al. Multiple output selection-LAS algorithm in large MIMO systems
Wang et al. Multiuser detection for uplink large-scale MIMO under one-bit quantization
Nguyen et al. Linear and deep neural network-based receivers for massive MIMO systems with one-bit ADCs
CN114051701A (en) Apparatus and method for machine learning assisted precoding
Satyanarayana et al. Multi-user hybrid beamforming relying on learning-aided link-adaptation for mmWave systems
CN107919895B (en) Distributed detection method of large-scale multi-user MIMO system
Kim et al. Semi-supervised learning detector for MU-MIMO systems with one-bit ADCs
Li et al. Practical interference exploitation precoding without symbol-by-symbol optimization: A block-level approach
CN110365414B (en) Enhanced optical space modulation method suitable for lognormal turbulence channel
CN113489510B (en) Large-scale MU-MIMO system DBP architecture signal detection method based on user grouping
Chataut et al. Huber fitting based ADMM detection for uplink 5G massive MIMO systems
Yao et al. High-efficiency expectation propagation detector for high-order massive MIMO systems
Li et al. Decentralized groupwise expectation propagation detector for uplink massive MU-MIMO systems
Dong et al. An enhanced fully decentralized detector for the uplink M-MIMO system
Yang et al. Over-the-air split learning with MIMO-based neural network and constellation-based activation
Castillo Soria et al. Multiuser MIMO downlink transmission using SSK and orthogonal Walsh codes
Jeon et al. A low complexity ML detection for uplink massive MIMO systems with one-bit ADCs
Beulah et al. Performance analysis of precoding techniques for Massive MU-MIMO systems
Seidel et al. A novel approach for combining local estimates for fully decentralized feedforward massive MIMO equalization: The multistep fusion
Ge et al. Approximate message passing MIMO detector with frequency domain initialization for SC-FDMA
Chen et al. Selection based list detection with approximate matrix inversion for large-scale MIMO systems
Seidel et al. Decentralized massive MIMO uplink signal estimation by binary multistep synthesis

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