.csma / Ca Mac Backoff Attack And Defense Research

Network Security is a fundamental prerequisite for network survivability and reliability in Ad hoc wireless network. However, the competitive and sharing wireless channel is vulnerable by Media Access Control (MAC) attacks from malicious nodes (or selfish nodes), which disobey the rules of random access backoff algorithms, for the purpose of the unfair occupation of channel resource, blocking channel or selfish benefit enhancement. There have been some detection methods in literature, but the existing strategy has the problem that calculation overhead is too heavy , the detection accuracy and the detection rate is too low, and can't cope with intelligent attack. This paper proposes a novel Central Limit Theorem (CLT) algorithm to detect MAC misbehaviors of malicious nodes, and design a detection protocol, CM-SEC(CLTMarkov-SECURITY) protocol, with the function of cooperating detection base on CLT algorithm.First, we have a introduce about the key technology in IEEE802.11 MAC (media access control ) protocol, summarize all sorts of attack behavior existing in MAC, have a deep study about the detection and defending methods coping with MAC attack, and abstract the excellence and fault. In order to conquer the most difficult backoff attack misbehavior, we propose a novel detect and defeat algorithm-CLT algorithm. For the characteristics of the binary exponential backoff, a single-step detection algorithm based on central limit theorem was presented in this paper, then for their less efficient problem, standardized questions was used to improve the treatment strategy, and complete theoretical analysis of detection performance. In order to improve the measurement accuracy of CLT single-step algorithm, the authors introduce the algorithm for Markov chain model, implement multi-step detection, and get the detection efficiency of the model in theory. Finally, simulation experiments show that the CLT algorithm is superior to the DOMINO algorithm, in terms of detection correctness and convergence speed, and verify that the CLT algorithm and Markov chain model of detection performance in theory is correct.Take advantage of CLT algorithm and Markov model, we present a corresponding detection protocol, CM-SEC protocol. Many tasks should be deal with during detection, such as information collection, data processing and farwording, so it is necessary to design specific detection protocol. The protocol works periodic for intelligent attack. Based on CLT algorithm and Markov model, we adopt the strategy of cooperating detection with multi-nodes, Based on CM-SEC protocol , we build a wireless sharing channel network model with MAC backoff attack with environment of Visual C++, Simulation experiments show that CM-SEC protocol can detect nodes of backoff attack accurately and efficiently, can improve the performance of network remarkably, and CM-SEC protocol just cause a little overhead in network.