Research On Wide-band Spectrum Sensing Algorithm For Aviation Radio Systems

Electromagnetic spectrum management is the foundation of aeronautical communications,which is the key to maximize the performance of avionics systems, an important guarantee forcommunication. At the conditions of modern information warfare, traditional static spectrummanagement and allocation patterns can not meet the demand. This paper intends to study awide-band spectrum sensing algorithm for aviation radio systems. We bring in the cognitiveradio technology to detect the electromagnetic environment that dynamic changed. Theresearch goals of this paper are reduce sample rate and computation complexity, lower the front-end hardware requirements, and improve spectrum detection speed. So that, we can utilizationthe radio spectrum resources more rational, and wish the proposed method can be applied tothe future aviation radio systems. The main work of this paper presents as following aspects.First, we introduce the research background of this paper, cognitive radio developmentprocess and some related research results. Summarizes some mature compressive sensingmethods and cooperative fusion algorithms, and analysis the pros and cons of differentalgorithms.Second, lucubrate the compressed sensing and sparse representation. Make a comparativestudy of sparse decomposition method from the sparse measure, sparse of function, thecomputing speed and precision of algorithm, and other aspects. Study of the sparsedecomposition method that based on frame-based and tracking algorithm, then figure out theneeded factors of accurately reconstruct the original signal with sparse decomposition algorithm.Then, research the broadband spectrum compression sensing algorithm that base on OMP.Introduce the OMP into broadband spectrum sensing. Calculating the energy of recoveryspectrum and combined with the results of each sensing node to determine the threshold andthat is used to determine the spectrum occupancy. Simulation results show that the algorithmcan make an acceptable reconstruction of the original signal in the circumstance of lowcompression ratio, and can maintain a high detection rate. At the same time, combine with the available cooperative spectrum sensing data fusionalgorithm, an improved WSPRT algorithm was proposed. It improves the reputation calculationalgorithm that based on the traditional WSPRT, and which makes identify malicious nodes moreprecise. While, except for the fusion weight of nodes, we add the recent sensing stability factor,which making the identification of malicious users more accurate and optimize system sensingperformance. Results of experiment show that the proposed algorithm reduces the influence ofSSDF nodes, improves the detection accuracy and stability of the system.