Reliable communication is a competitive task for broadband vehicular communications due to the fact that multiform interference has been introduced to the existing broadband transmission, which promotes the development of cognitive vehicular communication systems. To facilitate the improvement of the anti-jamming performance for the coexistence of diverse interference and signals in wireless heterogeneous networks, separating and eliminating various interference to cognitive communication systems is of importance. This paper formulated a novel sparse learning method-based cognitive transformation framework of interference separation for precise interference recovery, which can be efficiently resolved by iteratively learning the prior probability distribution of the sparse interference support. To further enhance the separation accuracy and iterative convergence, the principal component analysis and Bayesian perspective in orthogonal base learning were exploited to singly recover the multiple interference and communication signals. Moreover, through different sparsity states of spectrum analysis, the proposed novel interference separation algorithm was applied to simultaneous separation based on state evolving of approximation message passing, which iteratively learned the belief propagation posteriors and shrank by iterative shrinkage threshold. Simulation results demonstrate that the proposed methods were effective in separating and recovering sparse diversities of interference to communication systems, thereby significantly outperformed the state-of-the-art methods.

由于现有的宽带传输中引入了多种形式的干扰，从而促进了认知车载通信系统的发展，可靠通信是宽带车载通信的一项竞争任务。为了提高无线异构网络中多种干扰和信号共存的抗干扰性能，分离和消除对认知通信系统的各种干扰具有重要意义。本文提出了一种新的基于稀疏学习方法的干扰分离认知变换框架，用于精确干扰恢复，通过迭代学习稀疏干扰支持的先验概率分布可以有效解决该问题。为了进一步提高分离精度和迭代收敛性，利用正交基学习中的主成分分析和贝叶斯视角来单独恢复多个干扰和通信信号。此外，通过频谱分析的不同稀疏状态，将所提出的新干扰分离算法应用于基于近似消息传递状态演化的同时分离，迭代学习置信传播后验并通过迭代收缩阈值进行收缩。仿真结果表明，所提出的方法在分离和恢复对通信系统的稀疏干扰方面是有效的，从而显着优于最先进的方法。利用正交基学习中的主成分分析和贝叶斯视角来单独恢复多个干扰和通信信号。此外，通过频谱分析的不同稀疏状态，将所提出的新干扰分离算法应用于基于近似消息传递状态演化的同时分离，迭代学习置信传播后验并通过迭代收缩阈值进行收缩。仿真结果表明，所提出的方法在分离和恢复对通信系统的稀疏干扰方面是有效的，从而显着优于最先进的方法。利用正交基学习中的主成分分析和贝叶斯视角来单独恢复多个干扰和通信信号。此外，通过频谱分析的不同稀疏状态，将所提出的新干扰分离算法应用于基于近似消息传递状态演化的同时分离，迭代学习置信传播后验并通过迭代收缩阈值进行收缩。仿真结果表明，所提出的方法在分离和恢复对通信系统的稀疏干扰方面是有效的，从而显着优于最先进的方法。通过频谱分析的不同稀疏状态，将所提出的新型干扰分离算法应用于基于近似消息传递状态演化的同时分离，迭代学习置信传播后验并通过迭代收缩阈值进行收缩。仿真结果表明，所提出的方法在分离和恢复对通信系统的稀疏干扰方面是有效的，从而显着优于最先进的方法。通过频谱分析的不同稀疏状态，将所提出的新型干扰分离算法应用于基于近似消息传递状态演化的同时分离，迭代学习置信传播后验并通过迭代收缩阈值进行收缩。仿真结果表明，所提出的方法在分离和恢复对通信系统的稀疏干扰方面是有效的，从而显着优于最先进的方法。