Design Of Monitoring System For Ultra Short Wave Wireless Network Based On Software Defined Radio

Ultra short wave Ad Hoc network having the advantage of fast networking and robust structure is widely used in civil and military communication. However, the inherent characteristics including the exposure of the channel, multi- hop routing and dynamic topology make Ad Hoc network face many kinds of wireless network security threats, such as information cheating, false routing, DDos, and so on. Therefore, research on the wireless network monitoring in order to detect intrusion behaviors is very necessary for Ad Hoc network. The application of wireless monitoring needs to solve the three problems, including protocol analyzing, wireless data capturing and data analyzing. This paper focus on the analyzing of 220 D protocol, data capturing of data based on software definded radio and instrusion detection. The main work of this paper consists of the following aspects:(1) The security of wireless Ad Hoc is analyzed at first, according to the characteristics of wireless Ad Hoc network. Then a distributed monitoring system suitable for wireless Ad Hoc networks and the researching method for the three key technologies are proposed.(2) MS-STD-188-220 D protocol is learned. Then, the protocol analyzing for 220D/IP network is proposed to build a RF module based experimental 220 D platform which uses FSK modulation and sends packets similar to the PDU of 220 D protocol.(3) A SDR receiver is built by using u SDR3500 RF and baseband reconfigurable platform. The architecture and link for FSK receiver are designed and verified using Simulink tool. Then, the bit file is downloaded to u SDR3500 to capture and demodulate FSK signal generated by experimental 220 D platform, so the communication stream is obtained.(4) A method for flooding intrusion detection is researched using Hidden Markov Models. The changes of UDP traffics are used as observations of HMM to build a HMM with three hidden states. Then, the training sequences and testing sequences generated by NS-2 simulator are applied to train and evaluate HMM, respectively.The approach for UDP-Flooding detection is verified and the performance of detection model is analyzed.