面向多用户动态频谱接入的认知序列设计

Cognitive radio (CR) has been largely developed in recently years, and has been standardized for commercial and military applications. However, CR technology still meets some unsolved problems when applying for practical applications. For example, the sequences with good autocorrelation property have been used for time/frequency synchronization, channel estimation etc. In CR scenario, since only partial spectrum bands are available for cognitive users, the sequence elements should be forced to zero in order to avoid the inference from the occupied spectrum bands. Under this constraint, the traditional sequences transform to a new class of sequences with different correlation properties, including the autocorrelation and the cross-correlation. In this proposal, we consider a novel sequence design of time/frequency synthesis in order to obtain a cognitive sequence set with robust correlation properties. Meanwhile, we also consider the joint optimization algorithm of aperiodic autocorrelation and peak to average power ratio, which can be used for time/frequency synchronization and channel estimation in CR systems. In addition, based on the design of cognitive sequences, we consider the cognitive sequence based multiuser cognitive radio networks for achieving the properties of low probability of interception, proactive anti-jamming and anti-multiuser interference.

随着认知无线电(Cognitive Radio, CR)近几年的科学研究，CR技术已经开始标准化工作，并向实用化方向发展。然而在实用化过程中仍然面临一些问题，比如当传统具有良好相关特性的序列用于CR系统时间频率同步时，动态空闲频谱对应的序列元素(频域)将被保留，而占用频谱对应的序列元素将被设置为零。因此，传统序列将转变为满足空闲频谱约束条件的认知序列，且不再满足传统序列的相关特性。针对上述理论问题，本课题拟采用时间频率序列综合的方法进行认知序列集合研究，探索在动态空闲频谱约束条件下具有稳健相关特性的认知序列集合设计问题，建立认知序列相关特性和峰均功率比联合优化方法，为CR系统时频同步、信道估计所需的训练序列提供设计思路。此外，本课题拟在认知序列理论设计和优化的基础上，重点讨论基于认知序列设计的多用户认知无线网络应用所需解决的理论问题。

随着认知无线电(Cognitive Radio, CR)近几年的科学研究，CR技术已经开始标准化工作，并向实用化方向发展。然而在实用化过程中仍然面临一些问题，比如当传统具有良好相关特性的序列用于CR系统时间频率同步时，动态空闲频谱对应的序列元素(频域)将被保留，而占用频谱对应的序列元素将被设置为零。因此，传统序列将转变为满足空闲频谱约束条件的认知序列，且不再满足传统序列的相关特性。针对上述理论问题，本课题拟采用时间频率序列综合的方法进行认知序列集合研究，探索在动态空闲频谱约束条件下具有稳健相关特性的认知序列集合设计问题，建立认知序列相关特性和峰均功率比联合优化方法，为CR系统时频同步、信道估计所需的训练序列提供设计思路。此外，本课题拟在认知序列理论设计和优化的基础上，重点讨论基于认知序列设计的多用户认知无线网络应用所需解决的理论问题。