基于认知驱动的变换域通信智能抗干扰方法

为满足未来无人系统通信智能抗干扰的实际需要,针对传统变换域通信系统(transform domain communication system, TDCS)自身开放性有限、干扰应对能力不足等问题,设计了基于认知引擎驱动的智能系统架构,并针对各认知引擎驱动子模块提出了3种改进方法，包括基于稀疏逼近的未知干扰处理、基于稀疏表示的变换学习干扰识别以及针对性的干扰变换稀疏分析方法。实验结果表明,识别子模块与传统的分类器相比,整体的干扰识别率提高了5.2%,并且可实现无监督的学习;同时,针对典型干扰的重构精度在90%以上,实现了不同干扰类型的最优变换处理,显著提高了系统的抗干扰性能,传输误码率逼近理想水平。

Anti-interference method with intelligence for transform domain communication based on cognitive-engine

In order to satisfy the actual requirements of intelligent anti-interference for unmanned aerial systems communication in the future, an intelligent system driven based on cognitive engine is designed to solve the limited openness and insufficient ability for traditional transform domain communication system (TDCS). Furthermore, three improving methods are proposed for cognitive engine-driven modules, including unknown interference approximation based on sparse approximation, transform learning interference classification based on sparse representation and target sparse analysis method for interference transformation. Simulation results show that, compared with the traditional classifier, the overall classification of interference in the classification module is increased by 5.2%, and it can be realized in an unsupervised learning. Meanwhile, the accuracy of traditional interference reconstruction is above 90%, and the optimal transformation analysis for different interference is achieved, which significantly improves the anti-interference performance of the system and makes the bit error rate of transmission close to the ideal level.