Cooperative beamforming design for physical-layer security of multi-hop MIMO communications

Abstract

The security issue is critically important for wireless communications, especially for multi-hop communications. In this paper, we propose a cooperative secrecy beamforming scheme for a multi-hop MIMO communication network, in which there is a single-antenna source, multiple multi-antenna relays and a singleantenna destination. Moreover, two types of eavesdroppers exist in the network, one of which has known channel state information (CSI), while the other not. To achieve security communications, in our work null-space beamforming and artificial noise beamforming are jointly optimized to improve the secrecy rate of the multi-hop network. In nature, the joint optimization is nonconvex and challenging. Exploiting its structure, the considered optimization problem is decoupled into a series of subproblems that can be efficiently solved based on convex optimization theory. Finally, some numerical experiment results are provided to assess the performance of the proposed secrecy beamforming design.

创新点

物理层安全在无线网络通信中占据重要地位,尤其是针对窃听可能性较大的多跳网络。基于此我们提议了一个协作的多跳MIMO通信网络下的安全波束赋性方案。假定网络中存在一个单天线的发射端,多个多天线的中继以及一个单天线的接收端, 同时还有两个不同类型的窃听者, 其中一类窃听者的信道信息已知, 而另一类窃听者信道信息未知。为了实现多跳 MIMO 网络的安全通信,我们采用迫零波束赋性算法和人工噪声发送方案来分别尽可能的降低被这两类窃听者窃听的信息量,从而达到提升多跳网络的安全容量的目标。通常来说,对多个基站进行联合波束赋形设计有一定的困难和挑战,因为优化问题常常是非凸的NP-hard 问题。 因此, 我们提议了一个基于凸优化理论的迭代波束赋形优化算法,通过将原始的非凸优化问题分解为一系列具有凸性质的子问题来迭代的设计每个基站的最优波束矩阵。仿真结果表明我们所提议的算法可以有效的提升系统的安全容量。