The Studies Of Security In Wireless Mesh Network

As the wireless communication and mobile computing technology developed, as well as the requirement of the bandwidth, close attention is paid for WMN (Wireless Mesh Network). Different from the AP access technology with the single trip, there is no center point in WMN. The nodes in WMN connect each other by multi-hop. Thus, WMN is of the character of fast establishing, high reliability and pretty good self-adaptability. Amendment is made for the standard IEEE802.11s, IEEE 802.15 and IEEE 802.16, it makes WMN put in use in the scenario from WLAN to WAN. However, at the beginning of designing WMN, the main purpose is fast establishment and expandability, security is not the emphasis. In wireless application environment, WMN will be faced with more security problem, such as eavesdropping, intercepting and injecting. When WMN is no longer experimental network and put in commercial use, how to increase the security of WMN is a urgent need.In the early background investigation stage, I made full assessment of the existed schema. Based the investigation and the character of WMN, a distributed signature management schema which increase the security greatly is brought forward. The self-organized identity authentication is implemented by the trust-value, which can make judgment about the credibility and authentication ability of the node. To prevent WMN from compromised node, compromised node detection is added to WMN. By adopting the pro-detecting technology, the compromised node will be located and excluded from the network before real damage is caused, and the security of legal user will be protected. The three kinds of security schema would defend WMN from the internal attack and external attack. The security of whole WMN is improved greatly. After advanced the security model, I do the simulating work of the model by OPNET, and then analyze and forecast the capability of the new model. I analyzed the situations of no compromised node, no security detection, DBIL threshold equate 0.2 and DBIL threshold equate 0.8. After analyzed the simulation result, the validity of compromised node detection is verified and the application in particular network is discussed.