CN107947838A - The mixed-beam figuration framework and encoder design of a kind of Hybrid connections - Google Patents

Abstract

The invention discloses a hybrid beamforming architecture based on a hybrid connection, an encoder design algorithm and an implementation device. First, based on the design of the unconstrained hybrid encoder based on the idea of continuous interference cancellation, the optimal hybrid coding matrix is designed; secondly, when the optimal hybrid coding matrix is known, the constrained hybrid coding is designed based on the sub-optimized constrained hybrid encoder matrix, the analog encoding matrix and digital encoding matrix of the entire system can be obtained. This algorithm can effectively guarantee the spectral efficiency of the system and improve the energy efficiency of the system while reducing the complexity of system hardware implementation.

Description

A Hybrid Beamforming Architecture and Encoder Design for Hybrid Connections

technical field

The invention relates to the technical field of wireless communication, in particular to the design of a hybrid beamforming architecture and an encoder algorithm in a massive MIMO system.

Background technique

In order to meet the sharp increase in system capacity requirements of 5G, some new wireless communication technologies have emerged. However, the improvement of system capacity mainly includes the following three aspects: improving spectral efficiency, increasing the number of cells, and increasing spectral bandwidth; one of the corresponding key technologies is massive MIMO technology, ultra-dense network (UDN), millimeter wave ( mmWave) communication. However, the above-mentioned different technologies are not completely independent, and they promote and complement each other. For example, the shorter wavelength of mmWave makes it possible to arrange a large number of antennas in a limited physical space; large-scale antenna arrays can provide greater antenna array gain through beamforming, thereby making up for the loss of mmWave during transmission. However, some new problems also arise in the process of combining the above technologies with each other. For mmWave massive MIMO systems, traditional digital beamforming technology requires the same number of radio frequency (RF) links as the number of antennas, which is unrealistic in terms of specific hardware implementation. Therefore, traditional digital beamforming techniques are no longer suitable for massive MIMO systems.

As one of the key technologies of massive MIMO systems, hybrid beamforming technology can effectively reduce the number of RF links, thereby reducing the complexity of implementing large-scale antenna arrays. At present, there are two main types of hybrid beamforming architectures for large-scale antenna arrays: fully connected architectures and partially connected architectures. For the fully connected architecture, each RF link is connected to all elements of the antenna array, which consumes a large number of phase shifters, and the implementation complexity is high, but the degree of freedom of the system analog end is high, so it can Provides higher spectral efficiency. For the partial connection architecture, each RF link is connected to all the array elements of the sub-array, and the number of phase shifters that needs to be consumed is small, and the complexity of the implementation is low, resulting in a relatively low degree of freedom on the analog side of the system. Therefore, Compared with the fully connected architecture, the spectral efficiency of the system will be lost.

Contents of the invention

The present invention proposes a mixed-connected hybrid beamforming architecture, an encoder design algorithm, and an implementation device. This architecture can effectively ensure the spectral efficiency of the system while reducing the complexity of system hardware implementation, and effectively improve the efficiency of the system. energy efficiency.

Concrete realization process of the present invention is as follows:

The proposed hybrid beamforming architecture of hybrid connections is shown in Fig. 1, in which each subarray is connected to multiple RF links, and each RF link is connected to all antennas of the corresponding subarray.

For the hybrid beamforming architecture in Figure 1, the design of the hybrid encoder is divided into two parts: the design of an unconstrained hybrid encoder based on the idea of continuous interference cancellation; and the design of a suboptimal constrained hybrid encoder.

In the first part, for the hybrid connection HBF architecture in Figure 1, it is assumed that the number of antennas at the transmitting end is N _t , the number of antennas at the receiving end is N _r , and the channel matrix is The total number of RF chains in the system is _NR = SD, where D represents the number of sub-arrays, S represents the number of RF chains connected to each sub-array, and the digital terminal code is The analog side is coded as is a diagonal matrix, is the sent data stream, n is independent Gaussian white noise satisfying n∈CN(0,σ ² ), the signal at the sending end can be expressed as:

Where F=F _R F _B , the channel capacity of the system can be expressed as:

Each row of the hybrid block coding matrix F corresponds to the coding of a sub-array, and each column corresponds to the coding of a data stream. Meanwhile, the rows of the F matrix are similar to the analog encoding matrix, and the columns are similar to the digital encoding matrix. In order to simplify the algorithm expression, we put forward two assumptions:

(1) The number of data streams flowing through the subarray is equal to the number of RF chains connected to the subarray;

(2) The same data stream is transmitted in adjacent sub-arrays (without affecting algorithm performance).

Based on the above assumptions, we consider that different data streams are transmitted by the same multiple sub-arrays, the mixed block coding matrix F can be written as F=[F ₁ ,F ₂ ,...,F _K ], thus the optimization problem of the mixed coding matrix An optimization problem that can be decomposed into multiple encoding submatrices. The spectral efficiency of the system can be expressed as

Then the optimal mixed coding matrix F can be obtained according to the flow in FIG. 2 .

The second part is to design the constrained mixed coding matrix when the optimal mixed coding matrix F is known. Assuming that the constraint mixed coding matrix is G, matrix G and matrix F have the same structure, so G can be expressed as G=[G ₁ ,G ₂ ,...,G _K ], so that

According to the above description, we know that the optimal constraint coding matrix can be obtained by the following formula

Through the approximation of the entire matrix, it can be obtained that the encoding matrix corresponding to each sub-matrix is also approximate, so that

Regardless of the non-zero elements in G _i,sub , we can further get

in and are the non-zero elements in G _i,sub and F _i,sub respectively. To simplify the algorithm description, we assume and According to the previous method of matrix division, it can be divided into two parts, and the and undergo extreme decomposition to get

in P and M respectively satisfy P ^H P = I _S and M ^H M = I _S . Define B=[B ₁ ,B ₂ ] and W=[W ₁ ,W ₂ ], so that

According to the above formula, we can get the analog and digital encoding matrix corresponding to subarray i, which can be expressed as

right normalized to get

By repeating the above process D times, the analog encoding matrix and digital encoding matrix of the entire system can be obtained.

Description of drawings

Figure 1 shows the hybrid beamforming architecture for hybrid connections.

Fig. 2 is a block diagram of an unconstrained hybrid coding algorithm based on continuous interference cancellation.

Fig. 3 is a device for implementing an encoder under a hybrid beamforming architecture with hybrid connections.

Claims (6)

1. a kind of mixed-beam figuration framework of Hybrid connections and the mixed-beam forming algorithm for being common to existing framework, the framework Hardware consumption and implementation complexity are characterized in that between full connection and part connection framework, framework is connected entirely and part connects Framework is its special shape, and the framework has more preferable energy efficiency.The algorithm characteristics are to be suitable for above-mentioned three at the same time Kind framework, the hybrid coding matrix of unconfinement is obtained according to the thought of successive interference cancellation, it is encoded by constraint and unconfinement Between approximation relation constrained approximation relation between unconfinement subarray hybrid coding submatrix, by being carried out to matrix Polar decomghtion obtains the hybrid analog-digital simulation of suboptimum and digital encoder matrix.

2. according to the method described in claim 1, it is characterized in that, the framework has flexible design freedom, specifically include： The number that each subarray sends data flow can flexibly be adjusted according to the business demand of system；Secondly, RF number of links phases With in the case of, the number for the phase shifter that Hybrid connections framework needs is between full connection and part connection framework.

3. according to the method described in claim 1, it is characterized in that, according to the corresponding hybrid coding matrix of Hybrid connections framework Feature, splits hybrid coding matrix, and encoder matrix optimization problem, which is decomposed into multiple coded sub-matrices optimizations, asks Topic, then obtains the hybrid coding matrix of unconfinement using the thought of successive interference cancellation.

4. according to the method described in claim 3, it is characterized in that, compiled according to the hybrid coding matrix of unconfinement and constraint mixing Approximation relation between code matrix, obtains between each subarray unconfinement hybrid coding matrix and constraint encoder matrix Approximation relation, then finally obtains the encoder matrix of constraint according to the polar decomghtion of matrix.

5. according to claim 1, the method described in 2,3,4, it is characterised in that above-mentioned algorithm is equally applicable to connection and part entirely The mixed-beam figuration framework of connection.

6. a kind of mixed-beam size enlargement apparatus of Hybrid connections suitable for extensive antenna system, including：

Digital precode module, the data flow for system to be sent are mapped to corresponding RF links；

Precoding module is simulated, for the data flow for flowing through RF links being mapped to different bays.

Sub-array antenna module, for sending the signal of radio-frequency head.