Research On P2P Network Virus Detection Technology Based On Artificial Immune

Against today’s trends of the virus, researching a new method of virusdetection system to detect new kind of virus and reduce the rate of false alarm isespecially important. Existing virus detection systems, mostly single networkcentralized processing while current common distributed detection system hassome problems such as single point of failure, complex structure or systemoverhead. Biological systems and computer systems which is from the prototypeof organism are inextricably bound links, and at the same time, natural biologicalimmune system has the strong ability to distinguish between "self" and "non-self".Therefore, artificial immune method can be applied to computer virus detection.Firstly, this paper studies and analyses the development, structure andcharacteristic of P2P network, and then summarizes the deficiency of existingvirus detection system of P2P network. Points out the research direction of virusdetection on P2P networks in further; the related problems of artificial immunesystem are discussed in detail, mainly the development of artificial immunealgorithm. The advantages and disadvantages of each algorithm is compared andanalyzed. It is pointed out that the application of artificial immune algorithm incomputer virus detection has certain research significance.Finally, on the basis of artificial immune algorithm and routing algorithm,aiming at the problem of high false alarm rate, large bandwidth in the existingsystem, this paper proposes an improved virus detection model based on artificialimmune for P2P network, and uses an information fusion algorithm based on animproved viral genetic fingerprint distributed storage strategy as fusion strategybetween nodes in order to achieve inter node gene sharing in network. Virusdetection technology in P2P network means all nodes cooperating with each otherfor virus detection under the characteristics of the P2P network. In the wholedetection system, each node can generate virus gene independently using a guidly virus extraction algorithm, detect virus in local; once a node generates a newvirus gene, using Rabin fingerprint algorithm to extract feature code and realizingthe network sharing through Chord protocol on peer to peer networks. Finally,theoretical study and simulation of the model is done to evaluate the detectionperformance of the proposed model, and then summarizes the superiority anddeficiency in this proposed model.