多用户认知协作无线网络中的中继选择与载波分配策略

Relay selection and subcarrier allocation strategy in multiuser cognitive and cooperative wireless networks

在一个基于多载波的多主用户对-多次用户对-多认知中继节点组成的多用户认知协作无线网络中,本文考虑次用户网络采用underlay频谱接入模式与主用户网络共享频谱,次用户中继节点采用译码转发的信号处理方式,在满足主用户对干扰温度限制的条件下,研究如何通过中继选择和载波分配以最大化次用户对的信道容量和。该问题首先被建模成为一个带复杂约束的非凸整数规划问题。通过问题等价转化,本文把两个二维0-1系数矩阵的联合优化问题转化为一个三维0-1系数矩阵的优化问题,继而原问题转化为0-1线性规划问题。因为变量之间互相耦合的特性,常规优化算法无法求解此问题,基于分支定界框架,本文设计了中继选择与载波分配最优算法。仿真结果表明,该算法的复杂度远低于穷举法(平均迭代次数少于2.5次)且总能取到全局最优性能。

The paper focuses on optimizing cooperative wireless networks where there are multiple primary users, multiplesecondary users and multiple secondary relay nodes. The secondary users share a set of subcarriers with the primary users inunderlay mode, and decode-and-forward strategy is used at the secondary relay nodes. The objective is to maximize the aggregatecapacity of secondary networks by joint relay selection and subcarrier assignment, subject to the interference temperatureconstraints for the primary users. The resulting optimization problem is a constrained nonconvex integer programming problem,which is hard to solve because it needs to jointly optimize two bi-dimensional binary-coefficient matrices. To address this challenge,we propose to transform the original problem by transforming the joint matrices optimization into optimizing a singlethree-dimensional coefficient matrix, which can be further transformed to a binary linear programming problem. Since traditionaloptimization techniques cannot be directly applied because of the coupled variables, we design a solution algorithm basedon branch and bound framework. Simulation results indicate that the algorithm can always achieve the global optimum with verylow computational complexity( less than 2.5 iterations on average).