认知OFDM无线电系统中分布式资源分配

本文针对多用户的OFDM认知无线电系统，提出了一种适合于混合业务的分布式资源分配新算法。该算法以最大化系统容量为目标，将资源分配问题建模为非凸优化问题，并通过拉格朗日对偶理论将原问题分解为若干个独立的子问题，通过对子问题的求解可以获得最优的子载波分配和功率分配。同时，根据认知用户业务分组中不同的业务类型授予其不同的权重因子，确保资源分配结果能够满足各认知用户的QoS。仿真结果表明，该算法不仅提高了系统容量，而且还保证了资源分配的公平性和用户的QoS，且算法复杂度不高

Distributed Resource Allocation in Cognitive OFDM System

In This paper, a new algorithm of distributed resource allocation in cognitive OFDM system is addressed, which is suitable to mixed service. In order to maximum the throughput of system, we model the allocation problem as a non-convex optimization problem which be solved by Lagrangian duality theory, in which the original problem be decomposed into several independent sub problems. Though solving the several sub problems, we can obtain the optimal allocation results including sub carrier and power. in the meanwhile, according to the different type of service packets, we provide different cognitive user with the corresponding weight factor. The simulation results demonstrate that the proposed algorithm not only improves the throughput of system but also satisfies the QoS of cognitive user, and have low complexity.