认知无线电深度感知理论研究

项目名称： 认知无线电深度感知理论研究

项目编号： No.61471061

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 无线电电子学、电信技术

项目作者： 李斌

作者单位： 北京邮电大学

项目金额： 85万元

中文摘要： 认知无线电有望缓解频谱资源的紧缺局势，从而为未来宽带无线通信的发展开辟了一条新道路。作为其核心功能之一,现有频谱感知技术却难以应对未来复杂应用场景中的非理想特性（如移动场景与时变衰落），致使其检测性能无法接受。显著区别于已有方案，本课题拟提出一种全新的频谱感知框架。新方案首次将授权用户-次要用户间链路状态信息纳入考虑；在检测授权用户状态的同时，对上述链路状态信息执行联合估计，因而可称其为深度感知。深度感知通过深入发掘链路特性，可有望提升复杂场景下的检测性能；同时，估计获得的链路状态信息亦可优化后续认知资源调度。围绕全新深度感知理论框架，本研究主要工作包括（1）建立通用新感知模型，以深入刻画动态链路信息；（2）设计一类灵活高效的深度感知算法，实现对授权用户状态与链路信息的联合贝叶斯估计；（3）考虑四种典型应用场景并评估测试新算法的实际性能。本研究将对未来CR广泛应用具有重要的理论与实践意义。

中文关键词： 频谱检测；深度感知；授权-次用户链路状态信息；授权用户定位；随机有限集

英文摘要： As a promise technique that may alleviate the spectrum scarcity, cognitive radios (CRs) provide an attractive solution for the developments of emerging high-speed wideband wireless communications. Unfortunately, as a hard-core of CRs, the existing spectrum sensing schemes fail still to deal with the non-ideal characteristics encountered in future more complex scenarios, e.g. the mobile scenes and the time-varying fading, leading to a significantly degraded sensing performance. In sharp contrast to traditional schemes, this project will design a novel spectrum sensing framework. The new scheme, for the first time, will take the primary-secondary link state information (PS-LSI) in full considerations, which will be estimated jointly except for detecting the occupancy state of primary users. For this reason, we may refer the new sensing scheme to as deep sensing (DS). By thoroughly exploiting the underlying property of PS-LSI, the DS scheme would remarkably enhance the sensing performance even in those adverse conditions. Meanwhile, as another gift of sensing process, the recovered PS-LSI can be used for subsequent CR operation optimizations. Being centered on the new DS framework, the main work of this project may include the following three aspects, i.e., (1) establish a general sensing model which can profoundly reflect the dynamic property of PS-LSI; (2) design a flexible and efficient deep sensing algorithm to estimate jointly the unknown PU state and the PS-LSI; (3) consider four typical practical scenarios and evaluate the sensing performance of our proposed DS scheme. It is considered this project may be of great significance to future widespread use of CRs both in theory and practice.

英文关键词： spectrum sensing;deep sensing;primary-secondary link state information;primary user location;random finite set