基于启发式智能进化算法的MIMO协作频谱感知技术研究

Aiming at improving the timeliness, robustness and reliability of cooperative spectrum sensing system where the cognitive users are equipped with multiple antennas, this project will investigate the new techniques of the cooperative spectrum sensing by using the heuristic intelligent evolutionary algorithms. The project mainly focuses on the optimal allocation of sensing parameters and the optimization of cooperative sensing mechanism, attemptting to solve the following three problems: first is how to optimally design the numerous sensing parameters in order to maximize the sensing efficiency; second is how to jointly design the entire sensing-transmission cooperative spectrum sensing mechanism to improve the efficiency and reliablity of global spectrum sensing; finally is how to propose efficient and fast algorithms based on the heuristic intelligent evolutionary algorithms to solve the non-convex and nonlinear joint optimization problems mentioned above. The novelty of this project lies in the following aspects: the optimal configuration of adaptively dynamic sensing parameters with multiple constraints and multiple objectives, the real-time reliable cooperative global detection strategy based on multiple-dimensional diversity technique as well as the efficient and fast evolutionary heuristic intelligent algorithms in solving joint optimization problems. We dedicate ourselves to exploring real-time reliable spectrum sensing techniques and then providing a systemic MIMO cooperative global spectrum sensing strategy. This research is an integration of cognitive radio and human intelligence involving multiple subjects, with the expectation of having some breakthrough and novelty in both the theory and applications aspects, therefore provides some technical support for the research of the future connitive radio.

本课题拟针对认知用户配置多天线(MIMO)的协作频谱感知系统，以实时鲁棒可靠的频谱感知为目标，研究基于启发式智能进化算法的协作频谱感知技术的新思路。课题重点围绕系统感知参数的优化配置和协作感知机制的优化设计，解决以下相关问题：一是如何对众多感知参数进行优化配置以提高感知效率;二是如何对感知传输一体化的协作机制进行优化设计来改善全局检测的时效性和可靠性;三是如何设计求解上述非线性非凸联合优化问题的高效快速智能进化算法。本课题创新点在于提出多目标多约束条件下感知参数动态自适应优化配置方案，基于多维度分集技术的实时鲁棒增强全局检测方案以及高效快速求解联合优化问题的启发式智能进化算法，以此探索实时鲁棒可靠的协作频谱感知策略，最终形成系统的MIMO协作频谱感知方案。该研究是认知和人工智能的有机结合，涉及学科交叉，以期在理论和应用两个层面上有所突破和创新，从而为未来认知无线电的研究提供有力的技术支持。

本项目按照原研究计划，研究了多天线(MIMO)的协作频谱感知系统基于启发式智能进化算法的协作频谱感知技术的新思路，重点围绕MIMO系统参数优化配置，协作频谱感知机制优化设计以及进化算法求解协作感知联合优化问题展开，针对相关研究问题，给出了提高感知效率的感知参数优化配置方案，设计了改善时效性和可靠性的协作频谱感知方案，提出了求解非凸非线性优化问题的高效智能进化算法，相关研究成果发表在SCI检索论文，详见结题报告。.在课题的创新性方面，本项目提出的基于启发式智能进化算法的协作频谱感知技术为解决认知无线电频谱感知问题提供了新的的求解思路，它具有简单通用，适应性广，灵活性大，鲁棒性好，可并行求解，并且具有自适应调节能力的优势，目前吸引了国内外研究者广泛的关注，该研究隶属认知和人工智能的有机结合，期望未来能够在理论和应用层面有所突破和创新，从而为未来认知无线电的研究提供技术支持。另外，本项目后续将基于智能进化算法资源优化分配算法在FPGA或DSP中的仿真和实现，以验证该算法应用于实际资源分配的可行性、时效性和鲁棒性，为下一步的应用于实际认知MIMO系统奠定基础。.在基金委的资助下，本项目取得了一系列成果。三年来，共完成SCI论文10篇，EI检索期刊论文4篇，EI会议检索论文6篇，申请专利5项，已授权1项，毕业硕士2名，在读博士生2名，在读硕士生3名。综上所述，本项目按计划圆满完成了研究任务、达到了预期目标。

内容获取失败，请点击重试