Study And Design Of Key Evaluation Technologies For WLAN Security

WLAN security issues have been criticised currently. Due to the openness of wireless signal, wireless spoofing attacks is easy to launch. Current security events occur frequently such as user privacy, property loses and internal network intrusion. As wireless attack and defense technology constantly evolving, scholars began to pay attention in evaluation technologies for WLAN security. The goal of it is to detect and eliminate security threats. It is conventional to use cryptography techniques, while a wireless node identity can be verified by security measures such as authentication and authorization, but because of its high overhead requirements, practical application is far from satisfactory.Because RSS features are difficult to counterfeit, do not rely on cryptography and accessible. Therefore, this paper use the RSS associated with each wireless node, research the key technologies of WLAN security evaluation and design an RSS-based WLAN security evaluation method. Main contents include the following:(1) In view of the current WLAN security situation under IEEE 802.11, we analyse the security issues, such as common attack methods, approaches and procedures.(2) By using clustering algorithm of RSS and significance test, we give the qualitative conclusion of the WLAN attacks. In addition, we propose clusters of preferred method for quantitative analysis of WLAN attack.(3) We establish the fingerprint database and research WLAN attack localization technology. Though WLAN security evaluation result, we use a set of location algorithms to locate the attackers, such as the Nearest Neighbor, Weighted K-Nearest Neighbor and Bayesian Network method.(4) We extract the RSS feature in WLAN test environment and use the proposed method for WLAN security evaluation. And then we analyse evaluation performance indicators, including threshold and sample impact, qualitative and quantitative conclusion, attack effect and so on.The proposed WLAN security evaluation method has the following characteristics:give qualitative conclusion of WLAN spoofing attack, give quantitative results of WLAN spoofing attack, locate and track multiple attackers. Actual experimental results show that the proposed method can give over 90% of quantitative results and qualitative conclusion of detection rate and precision. Location results show the positioning of multiple attackers with good results.This research innovations mainly reflected in the following:(1) We design a WLAN security evaluation method and use the RSS feature and clustering algorithm. It can not only give quantitative results and qualitative conclusion for WLAN security, the results of quantitative analysis can also be used to accurately locate multiple attackers.(2) We use 3σ Criterion or Geluobusi Guideline for handling errors and use filters or Gaussian filter to filter the RSS, it can effectively reduce the impact of numerous wireless signal interference. Though combined with significant testing, the proposed method can further improve the detection rate of qualitative analysis for WLAN security evaluation.(3) The proposed cluster analysis algorithm can effectively overcome the strange cluster distributions and increase detection rate and precision on the WLAN security evaluation.(4) An attacker may send packets using different transmit power problems. By deploying multiple landmark, the proposed method can still detect and locate the location of the attacker.