多天线认知无线电网络自适应空间映射频谱共享

摘要

针对集中式多用户多天线认知无线电网络，提出一种基于自适应空间映射的频谱共享策略，根据授权用户接入的随机性所带来的频谱和空间资源占用情况的变化，认知系统将发射信号自适应地映射在认知基站与授权用户之间信道的子空间上，避免或抑制系统间干扰，从而在保证授权用户通信质量的前提下，为认知用户提供通信机会，并且在认知系统内部利用块对角化和奇异值分解方法分离不同认知用户的信号，消除系统内干扰。性能分析和仿真结果表明，与已有的频谱共享方法相比较，该策略不仅具有更高的认知系统可达和速率，并且对于由不同授权系统负载情况形成的不同场景具有更强的鲁棒性。%A spectrum sharing method based on adaptive spatial projection is proposed for central multi-user multi-antenna cognitive radio networks. In order to guarantee the transmission of primary users (PUs)as well as provide transmission opportunities to cognitive users(CUs),signals from cognitive base station(CBS)are adaptively projected onto the sub-spaces of the channel between CBS and PUs to avoid or reduce the interference between two systems according to the variances of spectrum and spatial resource occupancy caused by the random access of PUs. Furthermore, block diagonalization and singular value decomposition of channel matrices are adopted to cancel the interference between CUs. Performance analysis and simulation results show that, compared with existing methods, the proposed method can improve both the achievable sum rate of CUs and the robustness in different load scenarios.