Low-complexity multi-user MIMO cognitive radio system beam forming method

Abstract

The invention discloses a low-complexity multi-user MIMO (Multiple Input Multiple Output) cognitive radio system beam forming design method based on methods such as an improved subgradient algorithm and a classical WMMSE (Weighted Minimum Mean Square Error) algorithm, and compared with the existing MIMO multi-user cognitive radio beam forming technology, the method provided by the invention has the advantages that: on the premise of ensuring the weighted sum rate (WSR) of secondary users, the beam forming design of the MIMO multi-user cognitive radio system is realized; compared with the prior art, the method has the advantages that the calculation complexity is greatly reduced, especially when the number of antennas is large, the method has obvious advantages in the aspect of complexity, and application of an actual system is easier.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and more specifically, to a low-complexity multi-user MIMO cognitive radio system beamforming design algorithm combining a weighted MMSE optimization algorithm and an improved subgradient (MSM) algorithm. Background Art

[0002] With the rapid development of wireless communications and the sharp increase in data traffic, powerful physical layer technologies are needed to improve the spectrum efficiency within limited frequency band resources. Cognitive radio (CR) is considered to be an effective technology to improve spectrum efficiency and energy efficiency, so it has been deeply studied in domestic and foreign literature in the past decade. The cognitive radio system allows the primary user (PU) and the secondary user (SU) to share the same spectrum resource, while ensuring that the interference caused by the secondary base station (SBS) to the primary user remains within an acceptable range, guaranteeing the quality of service (QoS) of the PU. In order to better guarantee this requirement in the cognitive radio system, multiple input multiple output (MIMO) technology is often used for beamforming to ensure that the signal is enhanced as much as possible in the direction of the desired receiver. Although the low-complexity beamforming design in MIMO systems has been widely studied in the literature, there are relatively few algorithms in CR systems. Therefore, the focus of this patent is to develop a low-complexity beamforming algorithm for MIMO cognitive radio systems. Summary of the invention

[0003] Aiming at the common structure of MIMO cognitive radio system with multiple primary users and multiple secondary users, the present invention proposes a low-complexity beamforming design method capable of maximizing the weighted sum rate (WSR) in the scenario.

[0004] The present invention is based on constructing a low-complexity multi-user MIMO cognitive radio system beamforming design method, comprising the following steps:

[0005] S1: Assume that the channel state matrices from the primary base station to the kth secondary user, from the primary base station to the lth primary user, from the secondary base station to the kth secondary user, and from the secondary base station to the lth primary user are all known;

[0006] S2: According to the rules of cognitive radio, the primary base station has priority in the use of spectrum. The primary base station first performs beamforming based on the channel environment, and then the secondary base station considers beamforming without causing a significant impact on the primary base station and primary users, constructing the WSR maximization problem under L interference constraints and total power constraints.

[0007] S3: For the WSR maximization problem constructed in S2, the interference constraint and power constraint of the problem are first merged into a quadratic constraint using the improved subgradient method (MSM), and the problem becomes a WSR maximization problem with only one quadratic constraint;

[0008] S4: The problem obtained in S3 can be transformed into a WSR maximization problem without constraints but with the noise term as an adaptive term;

[0009] S5: Solve the problem obtained in S4 using the classic weighted MMSE algorithm, and use the block coordinate ascent method to solve the three sub-problems obtained by decomposition, and finally obtain the required beamforming matrix F k ,k=1,2,3,…K.

[0010] Among them, the present invention considers Figure 2 A MIMO cognitive radio system with L primary users and K secondary users, where both the primary base station and the secondary base station are equipped with N t Each primary user and each secondary user is equipped with N transmitting antennas. r The base station needs to transmit N receiving antennas to each user. s data streams. The channel state matrices from the primary base station to the kth secondary user, the primary base station to the lth primary user, the secondary base station to the kth secondary user, and the secondary base station to the lth primary user are G p,k ,H p,l ,H k ,G l , k=1,2,…K,l=1,2,…L. And the information vector transmitted by the primary base station for the lth primary user and the secondary base station for the kth secondary user is s p,l and k The beamforming matrices used are F p,l and F k .

[0011] Among them, the mathematical form of the optimization problem constructed by S2 is:

[0012]

[0013] Here, F={F1,F2,…F K},R k is the rate of the kth user, defined as

[0014]

[0015] α k is the weight of the kth user, representing its priority in the WSR calculation, P0 is the total power budget of the secondary base station, I l is the maximum interference power that the lth primary user can tolerate, is the noise variance of the kth user receiver, Represents N r dimensional identity matrix. To simplify the subsequent derivation, we set: Q: = {Q1, Q2, ..., Q K},C:={C1,C2,…C K}, and write the rate function as R k (Q,C).

[0016] Among them, S3 uses the improved subgradient method, specifically introducing the dual variable λ l ,l=0,1,2,…L, let The interference constraint and power constraint in the optimization problem constructed by S2 are combined to transform the optimization problem constructed by S2 into the following problem:

[0017]

[0018] Among them, S4 transforms the problem into the following form:

[0019]

[0020] Among them, S5 introduces two auxiliary variables Ω={Ω1,Ω2,…Ω K} and W={W1,W2,…W K}, and finally use the famous WMMSE conversion to transform the problem obtained by S4 into a weighted MMSE problem:

[0021]

[0022] Here E k Defined as Since this problem is convex for the three design variables, the theoretical optimal solution can be directly solved by the block coordinate descent method. According to the block coordinate descent method, the update rules of the three design variables are as follows:

[0023]

[0024] Here, Represents N s dimensional identity matrix.

[0025] According to the update rules, the above three variables are updated alternately, and the final converged design variable F k ,k=1,2,…,K is the required beamforming matrix. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 It is a diagram of the theoretical derivation of the present invention;

[0027] Figure 2 It is a schematic diagram of a scene model targeted by the beamforming design method proposed in the present invention;

[0028] Figure 3 It is a deployment location diagram of a primary user, a secondary user, a primary base station, and a secondary base station in a simulation experiment conducted by the present invention;

[0029] Figure 4 It is a curve showing the WSR performance of the algorithm proposed in the present invention and other comparison algorithms under the same conditions as the interference constraint threshold changes;

[0030] Figure 5 It is a curve showing the calculation time of the algorithm proposed in the present invention and other comparison algorithms changing with the number of secondary users under the same conditions. DETAILED DESCRIPTION

[0031] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0032] Example 1

[0033] use Figure 1 The beamforming design algorithm of the process shown in the figure is used for beamforming design (this method is named MSM-R-WMMSE algorithm), and the simulation results of the new method are compared with the results and calculation time obtained by the classic CVX solver and MOSEK solver, which includes the following steps:

[0034] S1: Assume that the channel state matrices from the primary base station to the kth secondary user, from the primary base station to the lth primary user, from the secondary base station to the kth secondary user, and from the secondary base station to the lth primary user are all known;

[0035] S2: According to the rules of cognitive radio, the primary base station has priority in the use of spectrum. The primary base station first performs beamforming based on the channel environment, and then the secondary base station considers beamforming without causing a significant impact on the primary base station and primary users, constructing the WSR maximization problem under L interference constraints and total power constraints.

[0036] S3: For the WSR maximization problem constructed in S2, the interference constraint and power constraint of the problem are first merged into a quadratic constraint using the improved subgradient method (MSM), and the problem becomes a WSR maximization problem with only one quadratic constraint;

[0037] S4: The problem obtained in S3 can be transformed into a WSR maximization problem without constraints but with the noise term as an adaptive term;

[0038] S5: Solve the problem obtained in S4 using the classic weighted MMSE algorithm, and use the block coordinate ascent method to solve the three sub-problems obtained by decomposition, and finally obtain the required beamforming matrix F k ,k=1,2,3,…K.

[0039] In the specific implementation process, the layout adopted by cognitive radio scenario simulation is as follows Figure 3 As shown, the channel is modeled using the Rayleigh channel

[0040] The WSR performance of the algorithm proposed in this invention is compared with the results obtained by the classic CVX solver. The results are as follows: Figure 4 As shown in the figure: no matter the number of secondary users is single or multiple, the MSM-R-WMMSE algorithm proposed in the present invention can achieve WSR performance consistent with other similar algorithms (including CVX-WMMSE algorithm based on CVX and Bis-WMMSE algorithm based on multiple dichotomies).

[0041] Embodiment 2:

[0042] The computational complexity is discussed based on Example 1. Specifically, in the process of setting simulation parameters, the interference constraint threshold is fixed to I l =10 -7 [W], the number of transmitting antennas is N t =64, we adjust the number of secondary users and record the calculation time of the beamforming matrix of each algorithm under different secondary users, so as to compare the computational complexity of different algorithms. The results are shown in Figure 5 As shown in the figure, it can be seen from the execution time of different algorithms under the same conditions that the MSM-R-WMMSE algorithm of the present invention requires the least solution time among all the optimization algorithms that can achieve the same WSR performance in this scenario, which means that the algorithm greatly reduces the computational complexity of the beamforming design problem in this scenario.

Claims

1. A low-complexity beamforming design method for a multi-user MIMO cognitive radio system, characterized in that: The following steps are included S1: Assume that the channel state matrices from the primary base station to the kth secondary user, from the primary base station to the lth primary user, from the secondary base station to the kth secondary user, and from the secondary base station to the lth primary user are all known; S2: According to the rules of cognitive radio, the primary base station has priority in the use of spectrum. The primary base station first performs beamforming based on the channel environment, and then the secondary base station considers beamforming without causing a significant impact on the primary base station and primary users, constructing the WSR maximization problem under L interference constraints and total power constraints. S3: For the WSR maximization problem constructed in S2, the interference constraint and power constraint of the problem are first merged into a quadratic constraint using the improved subgradient method (MSM), and the problem becomes a WSR maximization problem with only one quadratic constraint; S4: The problem obtained in S3 can be transformed into a WSR maximization problem without constraints but with the noise term as an adaptive term; S5: The problem obtained in S4 is solved using the classic weighted MMSE algorithm, and the three sub-problems obtained by the block coordinate ascent method are solved to finally obtain the required beamforming matrix.

2. The algorithm according to claim 1, characterized in that Block coordinate descent uses the following update rule:

Images