CN113452427B - Multi-cell cooperative large-scale MIMO robust precoding design and distributed processing method - Google Patents
Multi-cell cooperative large-scale MIMO robust precoding design and distributed processing method Download PDFInfo
- Publication number
- CN113452427B CN113452427B CN202110999819.3A CN202110999819A CN113452427B CN 113452427 B CN113452427 B CN 113452427B CN 202110999819 A CN202110999819 A CN 202110999819A CN 113452427 B CN113452427 B CN 113452427B
- Authority
- CN
- China
- Prior art keywords
- covariance
- base station
- cell
- channel
- users
- 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
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
- H04B7/0456—Selection of precoding matrices or codebooks, e.g. using matrices antenna weighting
Abstract
The invention discloses a multi-cell cooperative large-scale MIMO robust precoding design and distributed processing method, which comprises the following steps: (1) establishing a posterior channel model of a user; (2) establishing an interference plus noise covariance model of a user and a covariance model of a received signal; (3) establishing a traversal rate model or a traversal rate upper bound model of a user; (4) and solving the problems of weighted traversal and rate maximization or the problems of weighted traversal and rate upper bound maximization to obtain a precoding iterative expression. In the robust precoding design, the influence of channel estimation errors, channel aging and channel correlation on precoding can be resisted by establishing a posterior channel model of a user, and the problem of weight and rate maximization is solved by modeling user interference including inter-cell interference and noise. Three distributed processing methods with different information interaction quantities are disclosed, and multi-cell cooperative large-scale MIMO robust precoding design is concretely realized.
Description
Technical Field
The invention relates to a multi-cell collaborative precoding design and distributed processing method, in particular to a multi-cell collaborative large-scale MIMO robust precoding design and distributed processing method.
Background
In a large-scale Multiple-Input Multiple-output (M-MIMO) technology, a large number of antennas are arranged in a base station, so that not only can Multiple users be served simultaneously on the same time-frequency resource, but also higher frequency spectrum efficiency and energy efficiency can be achieved, and the technology becomes a key technology of a 5G physical layer. Through cooperation among base stations, the multi-cell technology can further improve the performance of a large-scale MIMO system. In the multi-cell cooperation technology, the same user can be simultaneously served by a plurality of base stations through interaction channel information and user data information among the base stations, and the rate performance of the user is greatly improved through spatial multiplexing. However, to receive the same user signal from different base stations results in more stringent synchronization requirements, and the interaction of user data between base stations significantly burdens the feedback channel. In contrast, in the multi-cell coordination technology, each base station independently serves users in the cell, and there is no overlap between cells. And each cell carries out cooperative precoding by avoiding the interference brought by other cells. The multi-cell cooperation technology not only reduces the signal processing difficulty, but also relieves the pressure of information interaction.
The large-scale MIMO performance gain is mainly achieved by downlink precoding. Linear precoding approaches the optimum performance with low complexity and is widely used in practice. In the research on massive MIMO linear precoding, most of them are based on accurate channel information, which is not available in practice. In a single-cell massive MIMO system, in order to effectively resist the influence of channel aging, channel estimation errors, and channel correlation, the related literature has studied how to perform robust precoding using an a posteriori channel model including a channel mean and a channel variance. Up to now, there is no research on multi-cell robust precoding based on the a posteriori channel model.
Disclosure of Invention
The invention provides a multi-cell cooperative large-scale MIMO robust precoding design. Another object of the present invention is to provide a distributed processing method based on the precoding design.
In order to achieve the purpose, the technical scheme of the invention is as follows: a multi-cell cooperative large-scale MIMO robust precoding design method comprises the following steps:
(1) establishing a posterior channel model of a user;
(2) establishing an interference plus noise covariance model of a user and a covariance model of a received signal;
(3) establishing a traversal rate model or a traversal rate upper bound model of a user;
(4) and solving the problems of weighted traversal and rate maximization or the problems of weighted traversal and rate upper bound maximization to obtain a precoding iterative expression.
Furthermore, each base station individually serves the users in the cell, and the cells are not overlapped. And each cell carries out cooperative precoding by avoiding the interference brought by other cells.
Further, the posterior channel model of the user contains channel mean and variance information.
Further, for a certain user, the interference plus noise covariance is equal to the sum of the covariance of noise, the covariance of equivalent interference channels in the cell, and the covariance of equivalent interference channels of other cells to the user; its received signal covariance is equal to the sum of the interference plus noise covariance and the covariance of the user's equivalent channel.
Further, in the case where the a posteriori channel information is known, the user's traversal rate or traversal rate upper bound is related only to user precoding and can be expressed as a function of the channel, user interference plus noise covariance, and precoding.
Further, the problem of weighted traversal and rate maximization precoding design can be solved by converting the MM algorithm into an iterative solution quadratic optimization problem; the precoding design problem of weighted traversal and rate upper bound maximization can be obtained by solving a KKT condition.
The invention relates to a multi-cell collaborative large-scale MIMO robust precoding whole-network parallel distributed processing method based on maximized weighted traversal and rate upper bound, which comprises the following steps:
(1) each base station acquires the posterior channel information of all users and generates initial precoding;
(2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(3) each base station interacts the covariance of equivalent interference channels of other cell users to the corresponding base station;
(4) for each user in the cell, each base station calculates the interference plus noise covariance and the covariance of received signals by using the covariance of the equivalent channel calculated in the step (2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained from the interaction of other base stations in the step (3);
(5) each base station broadcasts the interference plus noise covariance and the received signal covariance of the served users to other base stations;
(6) each base station updates the precoding of the served user by using the posterior channel information of all users, the interference plus noise covariance of all users and the received signal covariance;
(7) repeating (2) - (6) until convergence or a specified number of iterations is reached.
Further, the distributed processing is a robust precoding design method based on maximizing weighted traversal and an upper rate bound.
Further, in each iteration update, the interaction of all base stations needs to be performed synchronously.
The invention relates to a multi-cell collaborative large-scale MIMO robust precoding whole-network alternative distributed processing method based on maximized weighted traversal and rate upper bound, which comprises the following steps:
(1) each base station carries out initialization calculation and interaction;
(2) the active base station updates the precoding of the served user by using the posterior channel information of the active base station to all users, the interference plus noise covariance of all users and the received signal covariance;
(3) the active base station calculates the equivalent channel covariance of the served users in the cell, the covariance of equivalent interference channels in the cell and the covariance of equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(4) for each user in each cell, the active base station calculates the interference plus noise covariance and the received signal covariance of the equivalent channel covariance calculated in the step (3), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the users in other cells;
(5) if the iteration times reach the upper limit of the iteration times of a given single base station, selecting a new active base station from the standby base stations, sending the interference plus noise covariance and the received signal covariance of all users to the new active base station by the active base station, and then switching to a standby state;
(6) repeating (2) - (5) until convergence or a specified number of iterations is reached.
Further, the distributed processing is a robust precoding design method based on maximizing weighted traversal and an upper rate bound.
Further, precoding update alternates between base stations. The base station updating the precoding is called the active base station and the other base stations are called the standby base stations. Within the maximum iteration times of a given single base station, only the active base station updates the precoding of the users in the cell, and no information interaction exists.
Further, the step (1) comprises:
(1.1) each base station acquires the posterior channel information of all users and generates initial precoding;
(1.2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(1.3) each base station interacts the equivalent interference channel covariance of other cell users to the corresponding base station;
(1.4) for each user in the cell, each base station calculates the interference-plus-noise covariance and the received signal covariance by using the covariance of the equivalent channel calculated in the step (1.2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained in the step (1.3) from the interaction of other base stations;
(1.5) the standby base station transmitting the interference plus noise covariance and the received signal covariance of the served users to the active base station;
the invention relates to a multi-cell collaborative large-scale MIMO robust precoding base station-by-base station parallel distributed processing method based on maximized weighted traversal and rate upper bound, which comprises the following steps:
(1) each base station carries out initialization calculation and interaction and initializes a virtual network;
(2) the host station in each virtual network updates the precoding of the served users by using the posterior channel information of all users, the interference plus noise covariance of all users and the received signal covariance;
(3) the host station in each virtual network calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(4) for each user in each virtual network, each main base station calculates the interference-plus-noise covariance and the received signal covariance by using the equivalent channel covariance calculated in the step (3), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channels of other cell users;
(5) repeating (2) - (4) until convergence or a specified number of iterations is reached.
Further, the distributed processing is a robust precoding design method based on maximizing weighted traversal and an upper rate bound.
Further, precoding update is performed in parallel between the base stations. Each base station is used as a main base station to form a corresponding virtual network, and the precoding of each base station is iteratively updated in the virtual network. The virtual network comprises virtual base stations with the same number as other base stations and corresponding virtual users except the main base station.
Further, the precoding of the virtual base station remains unchanged throughout the precoding update process.
Further, the step (1) comprises:
(1.1) each base station acquires the posterior channel information of all users and generates initial precoding;
(1.2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(1.3) each base station interacts the equivalent interference channel covariance of other cell users to the corresponding base station;
(1.4) for each user in the cell, each base station calculates the interference-plus-noise covariance and the received signal covariance by using the covariance of the equivalent channel calculated in the step (1.2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained in the step (1.3) from the interaction of other base stations;
(1.5) each base station broadcasting the interference plus noise covariance and the received signal covariance of the served users to other base stations;
and (1.6) initializing the interference-plus-noise covariance and the received signal covariance of each virtual base station service user in each virtual network by utilizing the interference-plus-noise covariance and the received signal covariance of other base station service users obtained by interaction in the step (1.5).
Compared with the prior art, the invention has the following advantages and beneficial effects: the multi-cell large-scale MIMO robust precoding design provided by the invention not only can effectively inhibit inter-cell interference, but also can solve the problem of universality of multi-cell large-scale MIMO to various typical mobile scenes, and obtains high spectral efficiency compared with single-cell large-scale MIMO robust precoding and multi-cell large-scale MIMO non-robust precoding (see fig. 5 and 6). The three distributed processing schemes provided by the invention have the same iteration times, the full-network parallel distributed processing method has the highest frequency spectrum efficiency, the full-network alternate distributed processing method has the lowest computation complexity, and the base station-by-base station parallel distributed processing method has the lowest interaction overhead, so that different distributed implementation methods are provided for the practical application of the multi-cell large-scale MIMO robust precoding design.
Drawings
FIG. 1 is a flow chart of a multi-cell massive MIMO robust precoding design method;
FIG. 2 is a multi-cell cooperative massive MIMO robust precoding whole network synchronization distributed processing method;
FIG. 3 is a base station-by-base station sequential distributed processing method for multi-cell cooperative massive MIMO robust precoding;
FIG. 4 is a multi-cell cooperative massive MIMO robust precoding base station-by-base station parallel distributed processing method;
FIG. 5 is a graph comparing multi-cell cooperative massive MIMO robust precoding with WMMSE precoding performance and rate performance;
fig. 6 is a graph comparing distributed multi-cell cooperative massive MIMO robust precoding with single-cell robust precoding and non-robust precoding and rate performance.
Detailed Description
The technical solution of the present invention is further described below with reference to the accompanying drawings and examples.
As shown in fig. 1, it shows a flowchart of multi-cell massive MIMO robust precoding design according to the present invention, including:
(1) establishing a posterior channel model of a user;
(2) establishing an interference plus noise covariance model of a user and a covariance model of a received signal;
(3) establishing a traversal rate model or a traversal rate upper bound model of a user;
(4) and solving the problems of weighted traversal and rate maximization or the problems of weighted traversal and rate upper bound maximization to obtain a precoding iterative expression.
As shown in fig. 2, it shows a multi-cell cooperative massive MIMO robust precoding whole-network parallel distributed processing method based on maximized weighted traversal and rate upper bound, which includes:
(1) each base station acquires the posterior channel information of all users and generates initial precoding;
(2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(3) each base station interacts the covariance of equivalent interference channels of other cell users to the corresponding base station;
(4) for each user in the cell, each base station calculates the interference plus noise covariance and the covariance of received signals by using the covariance of the equivalent channel calculated in the step (2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained from the interaction of other base stations in the step (3);
(5) each base station broadcasts the interference plus noise covariance and the received signal covariance of the served users to other base stations;
(6) each base station updates the precoding of the served user by using the posterior channel information of all users, the interference plus noise covariance of all users and the received signal covariance;
(7) repeating (2) - (6) until convergence or a specified number of iterations is reached.
As shown in fig. 3, it shows a multi-cell cooperative massive MIMO robust precoding whole-network alternating distributed processing method based on maximum weighted traversal and rate upper bound, which includes:
(1.1) each base station acquires the posterior channel information of all users and generates initial precoding;
(1.2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(1.3) each base station interacts the equivalent interference channel covariance of other cell users to the corresponding base station;
(1.4) for each user in the cell, each base station calculates the interference-plus-noise covariance and the received signal covariance by using the covariance of the equivalent channel calculated in the step (1.2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained in the step (1.3) from the interaction of other base stations;
(1.5) the standby base station transmitting the interference plus noise covariance and the received signal covariance of the served users to the active base station;
(2) the active base station updates the precoding of the served user by using the posterior channel information of the active base station to all users, the interference plus noise covariance of all users and the received signal covariance;
(3) the active base station calculates the equivalent channel covariance of the served users in the cell, the covariance of equivalent interference channels in the cell and the covariance of equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(4) for each user in each cell, the active base station calculates the interference plus noise covariance and the received signal covariance of the equivalent channel covariance calculated in the step (3), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the users in other cells;
(5) if the iteration times reach the upper limit of the iteration times of a given single base station, selecting a new active base station from the standby base stations, sending the interference plus noise covariance and the received signal covariance of all users to the new active base station by the active base station, and then switching to a standby state;
(6) repeating (2) - (5) until convergence or a specified number of iterations is reached.
As shown in fig. 4, it shows a base station-by-base station parallel distributed processing method for multi-cell cooperative massive MIMO robust precoding based on maximized weighted traversal and rate upper bound, which includes:
(1.1) each base station acquires the posterior channel information of the base station to all users and generates initial precoding;
(1.2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(1.3) each base station interacts the equivalent interference channel covariance of other cell users to the corresponding base station;
(1.4) for each user in the cell, each base station calculates the interference-plus-noise covariance and the received signal covariance by using the covariance of the equivalent channel calculated in the step (1.2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained in the step (1.3) from the interaction of other base stations;
(1.5) each base station broadcasting the interference plus noise covariance and the received signal covariance of the served users to other base stations;
and (1.6) initializing the interference-plus-noise covariance and the received signal covariance of each virtual base station service user in each virtual network by utilizing the interference-plus-noise covariance and the received signal covariance of other base station service users obtained by interaction in the step (1.5).
(2) The host station in each virtual network updates the precoding of the served users by using the posterior channel information of all users, the interference plus noise covariance of all users and the received signal covariance;
(3) the host station in each virtual network calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(4) for each user in each virtual network, each main base station calculates the interference-plus-noise covariance and the received signal covariance by using the equivalent channel covariance calculated in the step (3), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channels of other cell users;
(5) repeating (2) - (4) until convergence or a specified number of iterations is reached.
The method is mainly suitable for a multi-cell large-scale MIMO system with a large-scale and super-large-scale antenna array arranged on a base station side to serve a plurality of users simultaneously. The following describes a specific implementation process of the method in detail with reference to a specific communication system example, and it should be noted that the method of the present invention is not only applicable to the specific system model exemplified in the following example, but also applicable to system models of other configurations.
First, system configuration
Consider a group of
Multiple cell system of cells operating in time divisionIn a duplex (TDD) mode, in which the base station
In a cell
Middle service
A user and is equipped with
A root antenna. By using
To indicate a cell
To (1)
And (4) users. For each base station
Consider a massive MIMO system equipped with a Uniform area Array (UPA). The uniform area array is common
Root antenna of which the vertical direction is
Root, horizontal direction
And (4) root. User' s
Is provided with a
A Uniform Linear Array (ULA) of root antennas. Assuming that the channel is approximately flat block fading, the system time resource is divided into a number of time slots, each time slot including
The channel remains unchanged over a time block. For simplicity, it is assumed that only uplink channel training and downlink transmission phases exist, and downlink transmission includes pre-coded field pilot and data signaling. In each time slot, the uplink pilot signal is transmitted only in the first time block. 2 to 2
The time block is used for transmitting the pilot frequency and the data signal of the downlink pre-coding domain. Each time slot obtains channel information for transmission of the time slot. For a Frequency Division Duplex (FDD) mode, the uplink channel training phase may be replaced with the downlink channel feedback phase, and the downlink transmission phase remains the same. Specifically, a downlink omni-directional pilot signal is transmitted in a first block, and mobile terminal feedback is received.
Two, multi-cell cooperative robust precoding
1. Posterior channel model
Cell
In the second of any time slot
The a priori channel over a time block may be expressed as
Wherein
Is a refined received sample steering vector matrix at the user side,
is a refined transmission sampling guide vector matrix at the base station side. Further comprises
Wherein
Is a refined guiding vector matrix in the vertical direction,
is a refined guide vector matrix in the horizontal direction,
to represent
Is accumulated and has
。
Refining the beam domain channel matrix for element independence, wherein
Representing a hadamard product. Each row corresponds to a refined beam domain on the user side, each column corresponds to a two-dimensional refined beam domain on the base station side,
in order to refine the beam domain channel magnitude matrix,
is a random matrix composed of independent identically distributed complex Gaussian random variables, the elements of which are zeroThe value unit variance. Defining a channel refinement beam domain energy matrix of a massive MIMO system as
Defining a refinement factor of
When the refinement factor is larger than 1, the number of cosine in the sampling direction is more than that of the antenna, and compared with the traditional wave beam domain prior statistical channel model based on the DFT matrix, the refined wave beam domain statistical model has more statistical characteristic directions, so that the actual physical channel model can be more accurately characterized.
Applying a first order Gaussian Markov process given channel information over a first time block, in the first time block
The beam domain channel over a time block can be modeled as
Wherein
Is that
And a velocity-dependent correlation coefficient. The a posteriori channel model can be expressed as
Wherein
The a posteriori channel (5) can depict a variety of actual mobility scenarios. Specifically, it degenerates to the channel estimation over the first time block
When is coming into contact with
Towards 1, where the channel may be approximated as a quasi-static channel. In the case of a high-mobility scenario,
approaching 0, the posterior channel model is similar to the prior channel model
。
Establishing an interference plus noise covariance model and a received signal covariance model of a user
Considering transmission on a single downlink time block, omitting time block corner marks
. Order to
Representing a user
The transmission of the data stream is performed,
representing a user
The number of data streams. The data symbol of the user is assumed to be a random matrix formed by independent and identically distributed complex Gaussian random variables, and the elements of the random matrix are zero-mean unitsThe variance. User' s
The received signal of (a) may be expressed as:
wherein
For the user
The precoding matrix of (a) is determined,
gaussian noise with zero mean unit variance, which is independent and identically distributed. The second term on the right of the equal sign of equation (8) is the intra-cell interference and the third term is the inter-cell interference. Adding noise to accumulated interference
As a Gaussian noise process, we define users
Has an interference plus noise covariance of
For the weighted traversal and rate upper bound maximization problem, defining users
Has a received signal covariance of
3. Establishing a traversal rate model or a traversal rate upper bound model of a user
Suppose that
At the user
To know, the user
The traversal rate of can be written as
Due to the fact that
Is a convex function and can obtain an upper bound rate of
4. Solving the problem of weighted traversal and rate maximization or the problem of weighted traversal and rate upper bound maximization to obtain a precoding iterative formula
4.1 weighted traversal and Rate maximization problem
Will be first
Precoding of individual base stations is represented as
The weighted traversal and rate maximization problem under the per-base-station power constraint can be written as
Wherein
Is a base station
Is determined. The objective function in the optimization problem (16) is an extremely complex function with respect to the precoding matrix, and therefore the problem is difficult to solve directly. The design problem of weighted traversal and rate Maximization precoding can be converted into an iterative solution quadratic optimization problem to be solved through an MM (minimization Maximization) algorithm. The key of the MM algorithm is to find a simple minorizing function of the objective function.
Defining functions
Wherein
And is
And
is irrelevant. Then
Is an objective function
the limit point of the precoding matrix sequence given in equation (23) is a local maximum point of the original optimization problem (16). Further, the optimization problem in equation (23) is the precoding matrix
A concave quadratic function of (a). The optimal solution can be directly obtained by a Lagrange multiplier method, namely
Wherein the content of the first and second substances,
and (4) an optimal Lagrangian factor corresponding to the energy constraint. For theMatrix array
Defining a single-sided correlation matrix
Wherein
The interference plus noise covariance is calculated as
Or
Wherein
Calculated from the following iterative Structure
the precoding iteration (24) becomes:
wherein:
in summary, the weighted traversal and rate-maximized robust precoding design can be summarized as the following steps:
step 1, setting
Randomly generating a set of precoding matrices
And normalizing it to meet the energy constraint;
step 2, calculation according to the formula (29)
Step 3, calculating according to the formulas (33), (34), (35), (36), (37) and (38)
Step 4, calculating according to the formulas (30), (39), (40), (41) and (43)
Repeating steps 2 through 5 until convergence or a predetermined target is reached.
4.2 weighted traversal and Rate Upper bound maximization problem
The problem of weighted traversal and rate ceiling maximization under per-base-station power constraints can be written as
The covariance of the user's received signal can be calculated as
Order to
The KKT condition of problem (44) can be written as
From the KKT condition, we construct a precoding iteration
Wherein
Can be calculated as
In summary, the weighted traversal and rate ceiling maximization robust precoding design can be summarized as the following steps:
step 1, setting
Randomly generating a set of precoding matrices
And normalizing it to meet the energy constraint;
step 2, calculation according to the formula (29)
；
Step 3, calculating according to the formula (45)
；
Step 4, calculating according to the formulas (30), (50) and (51)
；
Repeating steps 2 through 5 until convergence or a predetermined target is reached.
Three-cell multi-cell cooperative large-scale MIMO robust precoding whole-network parallel distributed processing method
For the user
Is pre-coded
Defining the covariance of the equivalent interference channel in the cell as
The other cell users have an equivalent interference channel covariance of
The interference plus noise of the user can be written as
Further, define the user
Has an equivalent channel covariance of
Its received signal covariance can be calculated as
The steps of the multi-cell collaborative large-scale MIMO robust precoding whole-network parallel distributed processing method based on the maximized weighted traversal and the upper rate bound can be summarized as the following steps:
step 1: is provided with
For each base station
Randomly generating a precoding matrix
And normalized to meet energy constraints;
step 2: for each base station
All are calculated according to equations (55), (52) and (53)
；
And step 3: for each base station
Each one of them
Interacting to the base station
And 4, step 4: for each base station
All are calculated according to equations (54), (56)
；
And 5: for each base station
All will be
Broadcast to other base stations;
step 6: for each base station
All are calculated according to equations (30), (50) and (51)
Updating the precoding according to equation (49)
Is provided with
。
Repeating steps 2 through 5 until convergence or a predetermined target is reached.
Four-cell and multi-cell cooperative large-scale MIMO robust precoding whole-network alternate distributed processing method
In the whole network alternative distributed processing method, only the active base station exists in each iteration
Updating its precoding, the standby base station
The interference plus noise covariance and the received signal covariance of the serving user may be individually based on
And
；
and (6) updating.
Giving maximum number of iterations of active base station
Defining the modulo operation as:
wherein
Is a rounding operation.
The steps of the multi-cell collaborative large-scale MIMO robust precoding whole-network alternating distributed processing method based on the maximized weighted traversal and the upper rate bound can be summarized as the following steps:
step 1: is provided with
For each base station
Randomly generating a precoding matrix
And normalizing it to meet the energy constraint;
step 2: for each base station
All are calculated according to equations (55), (52) and (53)
；
And step 3: for each base station
Each one of them
Interacting to the base station
；
And 4, step 4: for each base station
All are calculated according to equations (54), (56)
；
And 5: is provided with
For each base station
All will be
Interacting to the base station
；
Step 6: for base station
All are calculated according to equations (30), (50) and (51)
Updating the precoding according to equation (49)
，
And 7: for base station
All are calculated according to equations (55), (52) and (53)
；
And 8: for base station
All are calculated according to equations (54), (56)
All are calculated according to equations (57) and (58)
；
And step 9: if it is not
For the base station
Of the own cell
Of other cells
Interacting to the base station
，
Order to
；
Repeating steps 6 to 9 until convergence or a predetermined target is reached.
Five-cell and multi-cell cooperative large-scale MIMO robust precoding base station-by-base station parallel distributed processing method
In the base station-by-base station parallel distributed processing method, each base station independently updates own precoding under the condition that the interference among given cells is not changed. From another perspective, each base stationThe virtual network works in a virtual network, and the virtual network comprises the same virtual base station and users served by the same virtual base station as the virtual base station in the actual network except that the virtual base station is the main base station. The precoding of the virtual base station remains constant all the time. Specifically, for any
Indicating a base station
Corresponding cell in virtual network
User of a service
In the first place
Precoding after sub-iteration, for arbitrary
We have
Thus, at the base station
In a corresponding virtual network, a virtual base station
For the user
Is satisfied with the covariance of the equivalent interference channel
Then, the interference-plus-noise covariance of the host station serving user may be calculated as
The received signal covariance can be calculated as
The interference plus noise covariance of the virtual base station serving user may be calculated as
The received signal covariance can be calculated as
The precoding iteration in each virtual network is updated to become
Wherein
In summary, the steps of the base station-by-base station parallel distributed processing method based on the multi-cell collaborative massive MIMO robust precoding with maximized weighted traversal and the upper rate bound can be summarized as follows:
step 1: is provided with
For each base station
Randomly generating a precoding matrix
And normalized to meet energy constraints;
step 2: for each base station
All are calculated according to equations (55), (52), (53)
And step 3: for each base station
Each one is to be
Interacting to the base station
And 4, step 4: for each base station
Calculate all according to equations (54), (56)
And 5: for each base station
All will be
Broadcast to other base stations;
step 6: for each base station
In the pair thereofIn the virtual network, for the user of its service, order
For the users served by the virtual base station, order
And 7: for each base station
All are calculated according to equations (65), (66), (67)
And updating precoding according to equation (63)
；
And 8: for each base station
All are calculated according to equations (55), (52), (53)
；
And step 9: for each base station
Calculate all according to equations (60), (61)
All are calculated according to equations (62), (63)
(ii) a Order to
；
Repeating steps 7 to 9 until convergence or a predetermined target is reached.
Sixth, effect of implementation
In order to make those skilled in the art better understand the scheme of the present invention, traversal and rate performance display and different distributed processing performance display of the massive MIMO robust precoding transmission using multi-cell cooperation in this embodiment under specific system configuration are given below. The system is configured with the base station number of
Each base station being equipped with
Massive MIMO system with antennas, in which
. The number of users served by each base station is
Each user being equipped with
The uniform linear array. For simplicity, the moving speed of all users is set to be the same. Refinement factors at the base station are set to
Refinement factor at the user is
. Noise power free
Obtaining the boltzmann constant in simulation
Temperature in Kelvin
。
Fig. 5 shows the comparison of Weighted Minimum Mean Square Error precoding (WMMSE), multi-cell cooperative Weighted traversal and rate-maximized robust precoding solved using MM algorithm, and heuristic (heuristics) multi-cell cooperative Weighted traversal and rate-maximized robust precoding solved by KKT condition with the user moving speed of 30, 60, and 120 km/hour respectively. In the figure, the ordinate represents the average sum rate of each base station, and the abscissa represents the base station transmission power; as can be seen from the figure, in addition to the weighted traversal and the rate upper bound maximization precoding in the 120 km/h scenario, the sum rate performance of both robust precoding is better than the WMMSE precoding using instantaneous channel information.
Fig. 6 shows the comparison of the rate performance and the user moving speed of 30 km/h, 60 km/h and the case of using non-cooperative regularized zero forcing (NCRZF), non-cooperative robust (NCR), network-parallel (NP), network-alternating (NA), and base station-by-base-parallel (IP) precoding. In the figure, the ordinate represents the average sum rate of each base station, and the abscissa represents the base station transmission power; as can be seen from the figure, except for the case of low transmission power in the 120 km/h scenario, the sum rate performance of the three distributed processing methods of multi-cell cooperative robust precoding is better than that of single-cell robust precoding.
It should be noted that the above-mentioned embodiments are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention, and all equivalent substitutions or substitutions made on the basis of the above-mentioned technical solutions belong to the scope of the present invention.
Claims (7)
1. The multi-cell cooperative large-scale MIMO robust precoding design method is characterized by comprising the following steps of:
(1) establishing a posterior channel model of a user;
(2) establishing an interference plus noise covariance model of a user and a covariance model of a received signal;
(3) establishing a traversal rate model or a traversal rate upper bound model of a user;
(4) solving the problems of weighted traversal and rate maximization or the problems of weighted traversal and rate upper bound maximization to obtain a precoding iterative expression;
each base station independently serves users in the cell, the cells are not overlapped, each cell carries out cooperative precoding by avoiding interference brought by other cells, and a posterior channel model of the user comprises channel mean value and variance information;
for a certain user, the interference plus noise covariance is equal to the sum of the covariance of noise, the covariance of an equivalent interference channel in a cell and the covariance of the equivalent interference channel of the user by other cells; the received signal covariance is equal to the sum of the interference plus noise covariance and the covariance of the user's equivalent channel;
in the case that a posteriori channel information is known, the user's traversal rate or the upper bound of the traversal rate is only related to user precoding and is expressed as a function of the channel, user interference plus noise covariance, and precoding; converting the weighted traversal and rate maximization precoding design problem into an iterative solution quadratic optimization problem through an MM algorithm for solving; the precoding design problem of weighted traversal and rate upper bound maximization is obtained by solving a KKT condition.
2. Based on a multi-cell collaborative large-scale MIMO robust precoding whole-network parallel distributed processing method with maximized weighted traversal and upper rate bound,
characterized in that the method comprises the following steps:
(1) each base station acquires the posterior channel information of all users and generates initial precoding;
(2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(3) each base station interacts the covariance of equivalent interference channels of other cell users to the corresponding base station;
(4) for each user in the cell, each base station calculates the interference plus noise covariance and the covariance of received signals by using the covariance of the equivalent channel calculated in the step (2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained from the interaction of other base stations in the step (3); the interference plus noise covariance is equal to the sum of the covariance of the noise, the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user by other cells;
(5) each base station broadcasts the interference plus noise covariance and the received signal covariance of the served users to other base stations;
(6) each base station updates the precoding of the served users using the method of claim 1;
(7) repeating (2) - (6) until convergence or a specified number of iterations is reached.
3. The multi-cell cooperative large-scale MIMO robust precoding whole-network parallel distributed processing method based on the maximized weighted traversal and the upper rate bound as claimed in claim 2, wherein the distributed processing is a robust precoding design method based on the maximized weighted traversal and the upper rate bound, and the interaction of all base stations needs to be performed synchronously in each iteration update.
4. The multi-cell collaborative large-scale MIMO robust precoding whole-network alternating distributed processing method based on the maximized weighted traversal and the upper rate bound is characterized by comprising the following steps of:
(1) each base station performs initialization calculation and interaction, including:
(1.1) each base station acquires the posterior channel information of the base station to all users and generates initial precoding;
(1.2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(1.3) each base station interacts the equivalent interference channel covariance of other cell users to the corresponding base station;
(1.4) for each user in the cell, each base station calculates the interference-plus-noise covariance and the received signal covariance by using the covariance of the equivalent channel calculated in the step (1.2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained in the step (1.3) from the interaction of other base stations; the interference plus noise covariance is equal to the sum of the covariance of the noise, the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user by other cells;
(1.5) the standby base station transmitting the interference plus noise covariance and the received signal covariance of the served users to the active base station;
(2) the active base station updates the precoding of the served users using the method of claim 1;
(3) the active base station calculates the equivalent channel covariance of the served users in the cell, the covariance of equivalent interference channels in the cell and the covariance of equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(4) for each user in each cell, the active base station calculates the interference plus noise covariance and the received signal covariance of the equivalent channel covariance calculated in the step (3), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the users in other cells;
(5) if the iteration times reach the upper limit of the iteration times of a given single base station, selecting a new active base station from the standby base stations, sending the interference plus noise covariance and the received signal covariance of all users to the new active base station by the active base station, and then switching to a standby state;
(6) repeating (2) - (5) until convergence or a specified number of iterations is reached.
5. The method as claimed in claim 4, wherein the distributed processing is a robust precoding design method based on the maximum weighted traversal and the upper rate bound, and precoding updating is performed continuously among base stations.
6. The method for processing the robust precoding of the multi-cell collaborative massive MIMO in a parallel and distributed mode base station by base station based on the maximized weighted traversal and the upper rate bound is characterized by comprising the following steps:
(1) each base station performs initialization calculation and interaction, and initializes a virtual network, including:
(1.1) each base station acquires the posterior channel information of the base station to all users and generates initial precoding;
(1.2) each base station calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of other cell users based on the posterior channel information of all the users and the precoding of the served users;
(1.3) each base station interacts the equivalent interference channel covariance of other cell users to the corresponding base station;
(1.4) for each user in the cell, each base station calculates the interference-plus-noise covariance and the received signal covariance by using the covariance of the equivalent channel calculated in the step (1.2), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user in other cells obtained in the step (1.3) from the interaction of other base stations; the interference plus noise covariance is equal to the sum of the covariance of the noise, the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channel of the user by other cells;
(1.5) each base station broadcasting the interference plus noise covariance and the received signal covariance of the served users to other base stations;
(1.6) initializing the interference-plus-noise covariance and the received signal covariance of each virtual base station service user in each virtual network by using the interference-plus-noise covariance and the received signal covariance of other base station service users obtained by interaction in the step (1.5);
(2) the host station in each virtual network updating the precoding of the served users using the method of claim 1;
(3) the host station in each virtual network calculates the equivalent channel covariance of the served users in the cell, the covariance of the equivalent interference channels in the cell and the covariance of the equivalent interference channels of the users in other cells based on the posterior channel information of the host station to all the users and the precoding of the served users;
(4) for each user in each virtual network, each main base station calculates the interference-plus-noise covariance and the received signal covariance by using the equivalent channel covariance calculated in the step (3), the covariance of the equivalent interference channel in the cell and the covariance of the equivalent interference channels of other cell users;
(5) repeating (2) - (4) until convergence or a specified number of iterations is reached.
7. The method as claimed in claim 6, wherein the distributed processing is a robust precoding design method based on the maximum weighted traversal and the upper rate bound, precoding updating is performed in parallel among base stations, each base station serves as a host station to form a corresponding virtual network, and precoding of each base station is iteratively updated in the virtual network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110999819.3A CN113452427B (en) | 2021-08-30 | 2021-08-30 | Multi-cell cooperative large-scale MIMO robust precoding design and distributed processing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110999819.3A CN113452427B (en) | 2021-08-30 | 2021-08-30 | Multi-cell cooperative large-scale MIMO robust precoding design and distributed processing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113452427A CN113452427A (en) | 2021-09-28 |
| CN113452427B true CN113452427B (en) | 2021-11-19 |
Family
ID=77818872
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110999819.3A Active CN113452427B (en) | 2021-08-30 | 2021-08-30 | Multi-cell cooperative large-scale MIMO robust precoding design and distributed processing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113452427B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114567358B (en) * | 2022-03-03 | 2023-04-07 | 东南大学 | Large-scale MIMO robust WMMSE precoder and deep learning design method thereof |
| CN115001547B (en) * | 2022-05-31 | 2024-02-23 | 东南大学 | Multi-cell cooperation large-scale MIMO robust precoding design method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1863241A2 (en) * | 2002-08-15 | 2007-12-05 | Kabushiki Kaisha Toshiba | Channel tracking and signal detection in MIMO systems |
| CN111162822A (en) * | 2019-12-05 | 2020-05-15 | 东南大学 | Method and system for acquiring large-scale MIMO beam domain statistical channel information |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4799293B2 (en) * | 2005-09-21 | 2011-10-26 | 三洋電機株式会社 | Wireless device |
| CN107370693B (en) * | 2017-08-07 | 2019-09-13 | 电子科技大学 | Multi-user channel estimation method under extensive mimo system and DP priori |
-
2021
- 2021-08-30 CN CN202110999819.3A patent/CN113452427B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1863241A2 (en) * | 2002-08-15 | 2007-12-05 | Kabushiki Kaisha Toshiba | Channel tracking and signal detection in MIMO systems |
| CN111162822A (en) * | 2019-12-05 | 2020-05-15 | 东南大学 | Method and system for acquiring large-scale MIMO beam domain statistical channel information |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113452427A (en) | 2021-09-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen et al. | Channel estimation for reconfigurable intelligent surface aided multi-user mmWave MIMO systems | |
| Pan et al. | Multicell MIMO communications relying on intelligent reflecting surfaces | |
| CN107483088B (en) | Large-scale MIMO robust precoding transmission method | |
| Yin et al. | A coordinated approach to channel estimation in large-scale multiple-antenna systems | |
| CN113452427B (en) | Multi-cell cooperative large-scale MIMO robust precoding design and distributed processing method | |
| Li et al. | A multi-cell MMSE detector for massive MIMO systems and new large system analysis | |
| You et al. | Network massive MIMO transmission over millimeter-wave and terahertz bands: Mobility enhancement and blockage mitigation | |
| CN108199987B (en) | Downlink channel estimation method in FDD large-scale MIMO system | |
| Han et al. | Coordinated multi-point transmission strategies for TDD systems with non-ideal channel reciprocity | |
| Buzzi et al. | RIS configuration, beamformer design, and power control in single-cell and multi-cell wireless networks | |
| CN116438825A (en) | Positioning and auto-calibration in wireless networks | |
| WO2020232461A2 (en) | Methods and apparatus for multi-domain conversion of high dimensional channel statistics | |
| CN102724683A (en) | Multi-cell coordinated distributed robust beamforming method | |
| CN105991219B (en) | Method and device for processing transmission signal | |
| Chaves et al. | User selection for massive MIMO under line-of-sight propagation | |
| Qiu et al. | Low-complexity precoding by exploiting spatial sparsity in massive MIMO systems | |
| Zhang et al. | Spectral efficiency of massive MIMO networks with pilot contamination and channel covariance matrix estimation | |
| Xu et al. | DNN-aided codebook based beamforming for FDD millimeter-wave massive MIMO systems under multipath | |
| Zhang et al. | Efficient beamforming training and limited feedback precoding for massive MIMO systems | |
| Han et al. | FDD massive MIMO without CSI feedback | |
| Pichapati et al. | Practical considerations in cluster design for co-ordinated multipoint (CoMP) systems | |
| Khalilsarai et al. | Uplink-downlink channel covariance transformations and precoding design for FDD massive MIMO | |
| Ji et al. | Multipath extraction based UL/DL channel estimation for FDD massive MIMO-OFDM systems | |
| Amadid et al. | Pilot sequence-based channel estimation in massive mimo wireless communication networks under strong pilot contamination | |
| Khansefid et al. | Performance bounds for massive MIMO uplink |
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 |