Real-time Analysis And Intrusion Detection Of A Model Hf Adaptive Communication System Network Signal

A certain type of adaptive shortwave communication system has been massively fitted out, which undertakes shortwave data transmission tasks. It plays an irreplaceable role for construction of the military’s telecommunications as a final back-up means of communication. However, compared with wired network and even common wireless network, there are disadvantages such as limited bandwidth, high BER, complex topological characteristics, unbalanced network load and so on. Therefore, it’s particularly important to monitor the shortwave network and defense against intrusion proactively. This paper focuses on three issues of parsing shortwave wireless network protocols, data reception and data analysis, and research work are respectively conducted on the GJB 2077A-2007 protocol, wireless data reception and intrusion detection, and simulate and implement some of the functions. The main work includes:First, according to the characteristics of short-wave network and its security requirements, we propose a scheme of shortwave intrusion detection system.Second, GJB 2077A-2007 protocol standard of short-adaptive communication systems is studied, and its analysis method is given. By using 8PSK as the physical layer modulation method, an experimental platform is built to simulate the transmission of the system’s data frames, and used as the object of follow-up intrusion detection.Third, based on uSDR software radio platform, the simulation and link design of 8PSK receiver are completed by using Simulink, and the reception and demodulation of the 8PSK signal are realized so that the radio communication data stream can be acquired.Fourth, intrusion detection method is realized by modeling a hidden Markov chain, and applying the filter-based principal component analysis feature selection algorithm and setting a threshold based on Six Sigma method. The shortwave network intrusion model of threat, normal, suspected intrusion is established and the intrusion detection experiments are carried out. The detection model is trained and evaluated by learning and assessing the data packet acquired by NS-2 simulation, and the resutls prove its effectiveness.