认知OFDM系统的一种公平有效的多用户资源分配方案

摘要

在可用子载波、干扰温度、次级用户功率预算及公平性等约束条件下,提出了一种适用于认知OFDM系统的公平有效的多用户资源分配方案.该方案将子载波和功率分配分为两个相继的步骤,以降低计算复杂度.第一步引入子载波分配的比例公平原则,保证每个用户都能满足一定的通信要求;第二步采用一种改进的快速迭代注水算法,通过选择适当的调整步长μ值来改善算法收敛性能,可以使各次级用户在给定子载波上快速注水.算法分析和仿真结果表明,与现有的贪婪分配+迭代注水方案相比,该方案能灵活地在频谱有效性和公平性之间取得良好的折中,而且大幅降低现有算法的计算开销.%The paper presents a fair and efficient radio resource allocation scheme for multiuser OFDM-based cognitive radio systems, constrained by available subcarriers, interference temperature, secondary power budget, fairness, etc. The scheme adopts two steps to greatly reduce its computation complexity. The first step is to use a proportionally fair algorithm for subcarrier allocation to meet each users communication needs, and the second one is to use an improved efficient iterative water-filling algorithm and the proposed algorithm for fast power allocation to maximize the sum capacity. The extensive simulation results and the complexity analysis show that the proposed scheme offers a beneficial fairness-versus-capacity tradeoff compared to the existing solution, while significantly reducing the computation complexity.