CN110012489A - Communication processing method for full-duplex MIMO cellular system under non-ideal channel - Google Patents

Abstract

The invention discloses a communication processing method of a full-duplex MIMO cellular system under a non-ideal channel. A full-duplex MIMO cellular system includes several uplink devices, several downlink devices and base stations. The uplink device works in half-duplex mode, and the uplink and downlink devices communicate with the base station via the uplink; the base station works in the full-duplex mode, receives information sent by the uplink device through multiple antennas, and sends information to the downlink device at the same frequency at the same time; Establish a joint resource optimization problem, and divide it into an uplink and downlink device pairing sub-problem and a power allocation sub-problem and solve them separately to obtain the optimal uplink and downlink device pairing and the average rate of uplink and downlink device pairing. Communication relationship between devices. The resource optimal allocation method proposed by the invention effectively improves the system rate, can resist the system performance degradation caused by channel estimation error, self-interference and co-channel interference, and has the advantage of effective communication under non-ideal conditions.

Description

Communication processing method for full-duplex MIMO cellular system under non-ideal channel

technical field

The present invention relates to the technical field of wireless communication, and considers optimizing transmission power resources in a cellular network under the condition of non-ideal channel estimation. The present invention maximizes the system and rate by combining the pairing of uplink and downlink equipment and the power optimization allocation method. Finally, using decomposition and gradient projection algorithms, the optimization problem obtained by modeling is solved to obtain the optimal power distribution and maximize the system speed.

Background technique

Full-duplex (FD) technology can realize simultaneous bidirectional transmission of data on the same frequency band. Compared with traditional half-duplex (HD) technology, it can significantly increase the spectral efficiency. Therefore, Has broader development prospects. Due to the simultaneous transmission and reception on the same frequency, the full-duplex technology faces a serious Self-Interference (SI) problem. A lot of work has proposed various self-interference cancellation techniques from the theoretical research level and the hardware experiment level. At present, these self-interference cancellation techniques have sufficient ability to suppress self-interference to a lower level to meet the communication requirements.

However, most of the early researches are based on ideal conditions, that is, the receiver knows the channel state information (CSI) or does not consider co-channel interference (Co-channel-Interference, CCI), residual self-interference (Residual-Self-Interference, RSI) etc. However, in practical situations, due to the instantaneous change of the channel and the limitation of the wireless device itself, the complete CSI is difficult to obtain, and it is also very difficult to completely eliminate the SI and CCI. The design of full-duplex systems under non-ideal conditions has attracted the attention of scholars to a certain extent. Document 1 (D. Kim, H. Ju, S. Park and D. Hong, Effect of channel estimation error on full-duplex two-way networks, IEEE Tran Vehicular Tech., vol. 62, pp. 4666-4672; namely, D. Kim, H. Ju, S. Park and D. Hong, The effect of channel estimation error on full-duplex bidirectional networks, IEEE Technology for Internet of Things, vol. 62, pp. 4666-4672.) By using maximum ratio combining and optimal The ergodic capacity of the bidirectional full-duplex system is studied by combining method, and the influence of channel estimation error is observed. Reference 2 (A.C.Cirik, Y.Rong, and Y.Hua, Achievable rates of full-duplex MIMO radios in fast fading channels with imperfect channelestimation, IEEE Trans.Signal Process., vol.62, no.15, pp.3874-3886 , Aug.2014; i.e. A.C.Cirik, Y.Rong, and Y.Hua, Achievable rates of full-duplex MIMO radios in fast-fading channels under incomplete channel estimation, IEEE Signal Processing, vol.62, no.15, pp .3874-3886, Aug. 2014.) Assuming perfect self-interference cancellation and the existence of channel estimation errors, the achievable rates of bidirectional full-duplex MIMO systems are studied. The transmit beamforming problem of full-duplex point-to-point MIMO system under the condition of non-ideal channel state information is described in Reference 3 (J. Zhang, O. Taghizadeh, and M. Haardt, Robust transmit beamforming design for full-duplex point-to- point MIMO systems, in Proc.Tenth International Symposium on Wireless Communication Systems 2013, pp.346-350, 2013; namely J. Zhang, O. Taghizadeh, and M. Haardt, Robust Transmission Beamforming Design for Full-Duplex Point-to-Point MIMO Systems, 10th International Symposium on Wireless Communication Systems 2013, pp.346-350, 2013.). All of the above studies only consider the case of single-user MIMO bidirectional channels. In cellular environments, the impact of non-ideal channel state information on FD MIMO multi-user systems has not been seen so far.

SUMMARY OF THE INVENTION

In order to solve the problems existing in the background technology, in view of the difficulty in obtaining accurate channel state information in the actual communication environment and the self-interference of the full-duplex system and the co-channel interference of the cellular system, the present invention proposes a full-duplex channel in a non-ideal channel. The communication processing method of MIMO cellular system, which combines uplink and downlink device pairing and power allocation to optimize communication resources, and solves the problems caused by conditions such as countering channel estimation error, residual self-interference and co-channel interference to full-duplex MIMO cellular multi-user systems. performance degradation technical issues,

The technical scheme of the present invention comprises the following steps:

S1, the system model establishment steps:

A full-duplex MIMO cellular system includes several uplink devices, several downlink devices and base stations.

The uplink device works in half-duplex mode, the uplink device communicates with the base station via the uplink, and sends information to the base station through its own multiple antennas;

The base station works in full-duplex mode, receives information sent by uplink devices through multiple antennas, and sends information to downlink devices on the same frequency at the same time;

Downlink equipment equipped with multiple antennas, communicates with the base station via the downlink, and receives information sent by the base station;

Both the uplink device and the downlink device are equipped with multiple antennas, and the uplink device and the downlink device respectively communicate with the base station through channels corresponding to the multiple antennas.

S2, the joint resource optimization problem modeling step: using the total rate of the uplink and downlink as the system rate, establish a joint resource optimization problem of joint uplink and downlink device pairing and power allocation that maximizes the system rate;

S3, the joint resource optimization problem decomposition step: the joint resource optimization problem is decomposed into two sub-problems, which are the uplink and downlink device pairing sub-problem and the power allocation sub-problem respectively;

S4. Steps for solving the joint resource optimization problem: the gradient projection algorithm is used to solve the power distribution sub-problem, and the Hungarian algorithm is used to solve the sub-problem of pairing of uplink and downlink devices, so as to obtain the optimal pairing of uplink and downlink devices and the average of pairing (i, j) of uplink and downlink devices. rate, so as to allocate and set the communication relationship and communication resources between the base station and the uplink and downlink devices.

The steps of establishing the S1 system model are as follows:

A base station communicates with K uplink devices and J downlink devices at the same time. The uplink devices and the downlink devices serve as nodes, and the uplink devices and the downlink devices are both communication devices, such as mobile phones. Both the uplink device and the downlink device are equipped with N antennas, as shown in Figure 2;

The full-duplex MIMO cellular system has the following interferences: residual self-interference (RSI) caused by simultaneous transmission and reception of information by base stations on the same frequency, co-channel interference (CCI) caused by uplink and downlink sharing the same frequency, and channel estimation errors.

The minimum mean square error MMSE is used to estimate the channel, and the estimated value of the channel state information is expressed as The corresponding channel estimation error is expressed as ΔH, so the actual channel is expressed as in and ΔH are irrelevant. The elements in the channel estimation error ΔH are the mean value of 0 and the variance of Circular symmetric complex Gaussian variable of . The channel matrices are expressed as The four channel matrices correspond to the self-interference channel of the base station, the uplink channel from the i-th uplink device to the base station, the downlink channel from the base station to the j-th downlink device, and the interference channel of the i-th uplink device to the j-th downlink device, That is, H ₀ represents the self-interference channel matrix generated by the base station sending and receiving data at the same time, represents the uplink channel matrix from the i-th uplink device to the base station, represents the downlink channel matrix from the base station to the j-th downlink device, and H _ij is the interference channel matrix of the uplink device i to the downlink device j; respectively represent the four channel matrices H ₀ , The estimated values of the channel state information of H _ij , ΔH ₀ , ΔH _ij represents the four channel matrices H ₀ , The respective channel estimation errors of H _ij .

The signal received by the base station and the signal received by the j-th downlink device are represented as follows:

Among them, y ₀ represents the signal received by the base station, and y _j represents the signal received by the j-th downlink device; and is the IID and unit power data transmitted in the uplink and downlink, and is the transmit beamforming filter function of uplink and downlink; and Respectively represent the uplink channel matrix from the i-th uplink device to the base station and the downlink channel matrix from the base station to the j-th downlink device, and satisfy 1≤i≤K, 1≤j≤J, o indicates the base station, i indicates the uplink equipment The ordinal number of , j represents the ordinal number of downlink devices, K represents the total number of uplink devices, and J represents the total number of downlink devices; and Respectively represent the uplink channel matrix from the i-th uplink device to the base station The channel state information estimate and channel estimation error of , and ΔH ₀ represent the channel state information estimation value and channel estimation error of the self-interfering channel matrix H ₀ of the base station, respectively, and Respectively represent the downlink channel matrix from the base station to the jth downlink device The channel state information estimate and channel estimation error of , and ΔH _ij represent the channel state information estimation value and channel estimation error of the interference channel matrix H _ij of uplink device i to downlink device j, respectively; ρ _j and γ _i represent the power coefficients of residual self-interference and co-channel interference, respectively, n ₀ and n _j represent the white Gaussian noise (AWGN) received by the base station and the j-th downlink device, respectively, Both n ₀ and _{n j} obey a Gaussian distribution with mean 0 and variance IN;

The noise interference covariance matrix is calculated as:

where C _i and C _j are the noise interference covariance matrices of the i-th uplink device and the j-th downlink device, respectively, and is the channel matrix The estimated error power of H ₀ , H _ij ; and are the multi-antenna signal amplification power matrices of the i-th uplink device and the j-th downlink device respectively, I _N is an N×N unit matrix, tr{ } represents the trace of the matrix, and H represents the conjugate transpose of the matrix;

Amplify power matrix for multi-antenna signal and Perform eigenvalue decomposition:

where U _i and U _j are the unitary matrices of the eigenvectors of the i-th uplink device and the j-th downlink device, and is the diagonal power allocation matrix for the i-th uplink device and the j-th downlink device.

The steps of modeling the joint resource optimization problem in the step S2 are as follows:

The noise interference covariance obeys the Gaussian distribution, and the average rate of the pairing (i, j) of the uplink and downlink devices is obtained as:

in, To calculate the expectation, it is calculated by multiplying the variable by the probability distribution of the variable;

Establish the following joint resource optimization objective function that maximizes the system rate:

Among them, a _i,j represents the pairing parameters of the uplink and downlink devices, if the i-th uplink device and the j-th downlink device are paired to share the same channel resources, then a _i,j =1, otherwise a _i,j =0; and is the transmission power constraint of uplink and downlink signal transmission.

The step S3 is a joint resource optimization problem decomposition step, which is divided into two sub-problems as follows:

The joint resource optimization problem is an NP-hard problem, which requires an exhaustive search for all possible pairs a _ij , and when the number of uplink and downlink devices is large, a huge amount of computation will be generated. The present invention specifically decomposes the original optimization problem into two sub-problems , which can better solve this technical problem.

1) Power distribution sub-problem

In the case that the pairing of uplink and downlink devices is determined, the following power allocation objective function is constructed:

in, is a randomly generated pairing parameter of upstream and downstream devices, represents the average rate of pairing (i, j) of uplink and downlink devices, and respectively represent the diagonal power allocation matrix of the i-th uplink device and the j-th downlink device;

2) Sub-problem of uplink and downlink device allocation

On the basis of obtaining the optimal power allocation, the following objective function of uplink and downlink device allocation is constructed:

Among them, a _i,j are the pairing parameters of the uplink and downlink devices, Represents the average rate of each uplink and downlink device pairing (i, j) after power allocation optimization.

The step S4 is a joint resource optimization problem solving step, which is specifically as follows:

S41. In the initialization step, randomly generate a set of uplink and downlink device pairings (i,j) and their uplink and downlink device pairing parameters a _i,j , and randomly generate and allocate the power of the uplink and downlink devices, that is, the i-th uplink device and the j-th downlink device Device's diagonal power distribution matrix and

S42. Use gradient projection algorithm to iteratively optimize and solve the power allocation objective function according to the data randomly generated in step S41, and obtain the power of the optimal uplink and downlink devices, that is, the diagonal power allocation matrix of the i-th uplink device and the j-th downlink device and So that the average rate of maximum uplink and downlink device pairing (i, j) is obtained;

S43. Repeat the above steps S41-S42 to solve the power allocation objective function of the uplink and downlink devices by using different randomly generated uplink and downlink device pairs, and use the Hungarian algorithm to obtain the optimal pairing of the uplink and downlink devices.

In the specific implementation, in each iteration of the Hungarian algorithm, the rates of the i-th uplink device and the j-th downlink device are recorded, and K×J rates are generated in one iteration, and the Hungarian algorithm is used to search for different combinations of uplink and downlink devices. When the system rate is selected, the maximum system rate is output. At this time, the corresponding combination of uplink and downlink devices is used as the optimal pairing method of uplink and downlink devices.

The invention avoids the problem that accurate channel state information is difficult to obtain in the actual communication environment, establishes an optimization problem of joint uplink and downlink device pairing and power allocation to maximize the system rate without obtaining accurate channel state information, and proposes An asymptotic algorithm based on decomposition and gradient projection is proposed to solve the problem, and it is verified that the channel estimation error has the greatest impact on the performance in different types of interference.

Compared with the prior art, the advantages of the present invention are mainly reflected in the following aspects:

In the system model established by the present invention, both the base station and the user use multiple antennas to send and receive information, which can effectively improve the spectrum utilization rate, and is more in line with the actual communication process.

Secondly, in the cellular network of the present invention, the user works in the half-duplex mode due to equipment limitations, and the base station works in the full-duplex mode, which further improves the spectrum efficiency.

Finally, the communication resource optimization processing scheme of joint uplink and downlink device pairing and power allocation proposed by the present invention, compared with the existing resource optimization allocation method, considers the channel estimation error and the influence of various interferences in the system, and can counteract the channel estimation error, automatic System performance degradation due to interference and co-channel interference has the advantage of communicating effectively under non-ideal conditions.

In addition, the present invention also provides a reference for other related problems in the same field, which can be extended and extended based on this, and has a very broad application prospect when applied to technical solutions of other algorithms in the same field.

In general, the communication resource optimization processing method proposed by the present invention effectively improves the system rate, can resist the system performance degradation caused by channel estimation error, self-interference and co-channel interference, and has the advantage of effective communication under non-ideal conditions, The use effect is excellent, and it has high use and promotion value.

Description of drawings

FIG. 1 is a flowchart of a communication processing method for a full-duplex MIMO cellular system under a non-ideal channel.

FIG. 2 is a schematic structural diagram of a full-duplex MIMO cellular system based on non-ideal channel estimation conditions.

FIG. 3 is a schematic diagram of the convergence performance result of the algorithm proposed by the present invention.

FIG. 4 is a schematic diagram illustrating the influence of different types of interference on system performance.

FIG. 5 is a comparison diagram of pairing modes of uplink and downlink devices under different channel estimation errors.

Detailed ways

The specific embodiments of the present invention will be further described in detail below in conjunction with the accompanying drawings of the embodiments, so as to make the technical solutions of the present invention easier to understand and grasp.

The technical solution of the present invention is further described below in conjunction with the embodiments of the present invention implemented according to the complete method and the accompanying drawings:

Figure 1 is a flow chart of the communication processing method algorithm of a full-duplex MIMO cellular system under a non-ideal channel. First, initialization is performed, including average user power allocation and random pairing of uplink and downlink device pairing; then at the beginning of the 0th iteration, set random power allocation matrix and At the k-th iteration, gradient calculation is performed for gradient projection, and then the power allocation matrix is updated; when the algorithm converges, the iteration is stopped to obtain the optimal power allocation matrix; finally, the Hungarian algorithm is used to obtain the optimal pairing method of uplink and downlink devices; At this point, the entire system and the rate are at their maximum, and the algorithm ends.

Fig. 2 is a communication processing method of a full-duplex MIMO cellular system under a non-ideal channel, considering that there are K uplink devices and J downlink devices in the system, and each node is equipped with N antennas. The base station works in full-duplex mode, and the user works in half-duplex mode due to equipment limitations. In the actual environment, channel estimation is not ideal, and there is a channel estimation error. At the same time, the base station contains residual self-interference (RSI), and the downlink contains co-channel interference (CCI).

FIG. 3 is a graph of the convergence performance of the algorithm proposed by the present invention. As shown in the figure, under different residual self-interference, the joint optimization algorithm achieves convergence in only a few iterations, and the system can achieve a higher rate with lower residual self-interference. It shows that the algorithm of the present invention is low in complexity and has broad application prospects.

Figure 4 shows the impact of different types of interference on a full-duplex MIMO cellular system. The solid line indicates that the uplink and downlink have the same transmit power, and the dashed line indicates that the downlink has ten times the transmit power of the uplink. As shown in the figure, when the uplink transmit power is equal to the downlink transmit power, the effects of RSI and CCI on the system rate are the same. At this time, the CSI has the greatest impact on the system rate, because the RSI only exists in the uplink, the CCI only exists in the downlink, and the CSI exists in both the uplink and the downlink. When the transmission power of the downlink is much higher than that of the uplink, the CCI is relatively low due to the low transmission power of the uplink, and the RSI is relatively high due to the high transmission power of the downlink, so the influence of the RSI on the total system rate becomes more obvious. In addition, CSI exists in both uplink and downlink, which still has the most serious impact on the system rate.

Figure 5 tests the importance of uplink and downlink device pairing in the optimization problem of a full-duplex MIMO cellular system. The figure shows that, compared with the random uplink and downlink device pairing scheme, the uplink and downlink device pairing scheme proposed by the present invention has higher CSI error , still exhibits lower CSI errors than random uplink and downlink device pairings better performance. At the same time, it is found that the total system rate decreases with the increase of channel estimation error and RSI.

To sum up, the present invention establishes an optimization problem of joint uplink and downlink device pairing and power allocation in view of the difficulty in obtaining accurate channel state information in the actual communication environment, the self-interference of full-duplex systems and the multi-user interference of cellular systems. To maximize the system rate, an asymptotic algorithm based on decomposition and gradient projection is proposed to solve the problem, and it is verified that the channel estimation error has the greatest impact on performance in different types of interference. The resource optimal allocation method proposed by the invention can effectively improve the system rate, can resist the system performance degradation caused by channel estimation error, self-interference and co-channel interference, and has the advantage of effective communication under non-ideal conditions.

In addition, the present invention also provides a reference for other related problems in the same field, which can be extended and extended based on this, and has a very broad application prospect when applied to technical solutions of other algorithms in the same field.

The communication processing method of the full-duplex MIMO cellular system under the non-ideal channel proposed by the present invention has excellent use effect and high use and promotion value.

Equivalent structural transformations of the present invention made by those skilled in the art according to the description and the accompanying drawings are all included in the patent scope of the present invention.

Claims (6)

1. the communication processing method of full-duplex MIMO cellular system under a non-ideal channel, is characterized in that, comprises the following steps: S1, the system model establishment steps: The uplink device works in half-duplex mode, the uplink device communicates with the base station through the uplink, and sends information to the base station through its own multiple antennas; the base station works in full-duplex mode, receives the information sent by the uplink device through multiple antennas, and at the same time Send information to downlink devices on the same frequency; downlink devices equipped with multiple antennas communicate with the base station via the downlink, and receive information from the base station; S2, the joint resource optimization problem modeling step: using the total rate of the uplink and downlink as the system rate, establish a joint resource optimization problem of joint uplink and downlink device pairing and power allocation that maximizes the system rate; S3, the joint resource optimization problem decomposition step: the joint resource optimization problem is decomposed into two sub-problems, which are the uplink and downlink device pairing sub-problem and the power allocation sub-problem respectively; S4. Steps for solving the joint resource optimization problem: the gradient projection algorithm is used to solve the power distribution sub-problem, and the Hungarian algorithm is used to solve the sub-problem of pairing of uplink and downlink devices, so as to obtain the optimal pairing of uplink and downlink devices and the average of pairing (i, j) of uplink and downlink devices. rate, so as to set the communication relationship between the base station and the uplink and downlink devices. 2. the communication processing method of full-duplex MIMO cellular system under a kind of non-ideal channel as claimed in claim 1, is characterized in that: described S1 system model establishment step, is specifically as follows: A base station communicates with K uplink devices and J downlink devices at the same time, and both uplink devices and downlink devices are equipped with N antennas; the signal received by the base station and the signal received by the jth downlink device are expressed as follows: Among them, y ₀ represents the signal received by the base station, and y _j represents the signal received by the j-th downlink device; and is the IID and unit power data transmitted in the uplink and downlink, and is the transmit beamforming filter function of uplink and downlink; and Respectively represent the uplink channel matrix from the i-th uplink device to the base station and the downlink channel matrix from the base station to the j-th downlink device, and satisfy 1≤i≤K, 1≤j≤J, o indicates the base station, i indicates the uplink equipment The ordinal number of , j represents the ordinal number of downlink devices, K represents the total number of uplink devices, and J represents the total number of downlink devices; and Respectively represent the uplink channel matrix from the i-th uplink device to the base station The channel state information estimate and channel estimation error of , and ΔH ₀ represent the channel state information estimation value and channel estimation error of the self-interfering channel matrix H ₀ of the base station, respectively, and Respectively represent the downlink channel matrix from the base station to the jth downlink device The channel state information estimate and channel estimation error of , and ΔH _ij represent the channel state information estimation value and channel estimation error of the interference channel matrix H _ij of uplink device i to downlink device j, respectively; ρ _j and γ _i represent the power coefficients of residual self-interference and co-channel interference, respectively, n ₀ and n _j represent the white Gaussian noise (AWGN) received by the base station and the j-th downlink device, respectively; The noise interference covariance matrix is calculated as: where C _i and C _j are the noise interference covariance matrices of the i-th uplink device and the j-th downlink device, respectively, and is the channel matrix The estimated error power of H ₀ , H _ij ; and are the multi-antenna signal amplification power matrices of the i-th uplink device and the j-th downlink device respectively, I _N is an N×N unit matrix, tr{ } represents the trace of the matrix, and H represents the conjugate transpose of the matrix; Amplify power matrix for multi-antenna signal and Perform eigenvalue decomposition: where U _i and U _j are the unitary matrices of the eigenvectors of the i-th uplink device and the j-th downlink device, and is the diagonal power allocation matrix for the i-th uplink device and the j-th downlink device. 3. The communication processing method of a full-duplex MIMO cellular system under a non-ideal channel as claimed in claim 1, wherein: the joint resource optimization problem modeling step described in the step S2 is as follows: The noise interference covariance obeys the Gaussian distribution, and the average rate of the pairing (i, j) of the uplink and downlink devices is obtained as: in, for calculating expectations; Establish the following joint resource optimization objective function that maximizes the system rate: Among them, a _i,j represents the pairing parameters of the uplink and downlink devices, if the i-th uplink device and the j-th downlink device are paired to share the same channel resources, then a _i,j =1, otherwise a _i,j =0; and is the transmission power constraint of uplink and downlink signal transmission. 4. the communication processing method of full-duplex MIMO cellular system under a kind of non-ideal channel as claimed in claim 1, is characterized in that: described step S3 joint resource optimization problem decomposition step, is divided into two sub-problems specifically as follows: 1) Power distribution sub-problem Construct the following power distribution objective function: in, is a randomly generated pairing parameter of upstream and downstream devices, represents the average rate of pairing (i, j) of uplink and downlink devices, and respectively represent the diagonal power allocation matrix of the i-th uplink device and the j-th downlink device; 2) Sub-problem of uplink and downlink device allocation Construct the following upstream and downstream device allocation objective functions: Among them, a _i,j are the pairing parameters of the uplink and downlink devices, Represents the average rate of each uplink and downlink device pairing (i, j) after power allocation optimization. 5. the communication processing method of full-duplex MIMO cellular system under a kind of non-ideal channel as claimed in claim 1, is characterized in that: described step S4 joint resource optimization problem solving step, is specifically as follows: S41. In the initialization step, randomly generate a set of uplink and downlink device pairings (i,j) and their uplink and downlink device pairing parameters a _i,j , and randomly generate and allocate the power of the uplink and downlink devices, that is, the i-th uplink device and the j-th downlink device Device's diagonal power distribution matrix and S42. Use gradient projection algorithm to iteratively optimize and solve the power allocation objective function according to the data randomly generated in step S41, and obtain the power of the optimal uplink and downlink devices, that is, the diagonal power allocation matrix of the i-th uplink device and the j-th downlink device and So that the average rate of maximum uplink and downlink device pairing (i, j) is obtained; S43. Repeat the above steps S41-S42 to solve the power allocation objective function of the uplink and downlink devices by using different randomly generated uplink and downlink device pairs, and use the Hungarian algorithm to obtain the optimal pairing of the uplink and downlink devices. 6. The communication processing method of a full-duplex MIMO cellular system under a non-ideal channel as claimed in claim 1, wherein the full-duplex MIMO cellular system has the following interference: the base station simultaneously sends and receives information on the same frequency to generate Residual self-interference (RSI), co-channel interference (CCI) caused by sharing the same frequency for uplink and downlink, and channel estimation error.