The Research Of Mobile Ad Hoc Network Clustering Algorithm

Mobile Ad Hoc network (MANET) is a special kind of self-organizing peer topeer multi-hop wireless mobile network. With the popularity of wireless networktechnology and application since1990s, Ad Hoc network is running in a networkenvironment of independence without a network infrastructure with its advantages inmany aspects, such as no center self-organizing, multiple hops routing, dynamictopology, etc. In practical applications, in order to get better network scalability, thehierarchy Ad Hoc network is widely favored. In order to increase the throughput andscalability of network and improve the network scale, each node forms differenthierarchies by clustering algorithm, and shares the function of different levels, thusthe clustering algorithm directly affects the performance of hierarchy Ad Hocnetwork.At present, the clustering algorithm for Ad Hoc network has a large number ofresearch results, mainly includes clustering algorithm based on a single parameter, theweighted clustering algorithm based on multi-parameter, and the adaptive clusteringalgorithm based on the multi-parameter, etc. The selections of parameters are mainlyconcentrated in the mobility, stability, energy of mobile nodes. However, the study ofthe security of nodes is less for mobile Ad Hoc networks which are more likely toappear in the process of clustering, and the selections of the parameters in weightedclustering algorithm have yet to be optimized further.This paper is a discussion and research on the parameter selection of clusteringalgorithm and safety performance for the hierarchical mobile Ad Hoc network. Itjoins security consideration on the basis of traditional weighted clustering algorithmand adopts different methods to determine the weighting parameters respectively,aims to improve the clustering performance of Ad Hoc network, at the same time,improve the security performance. The main work of this paper includes:(1) In-depth analyzed the existing clustering algorithms of hierarchical mobileAd Hoc network, summarized each principle and technology, and compared the advantages and disadvantages of the algorithms. Analyze the security problems whichexisting algorithms are still facing and propose the solutions of this article.(2) According to the security problems that may be encountered in the clusteringof Ad Hoc network, proposed an adaptive Security Weighted Clustering Algorithm(SWCA) which adds security measure parameter. In addition to considering thedegrees of nodes, relative mobility, residual energy value, the safety parameterscalculation is joined, which includes the internal and external security parameters toensure its accuracy, so as to identify network multiple attacks, improve the securityperformance of Ad Hoc network clustering.(3) Given the required data structure, determined the process of calculatingparameters, and discussed cluster management process including form a clustering,join and exit the clustering based on the above algorithm. Eventually multiple sets ofcontrast experiments used network simulation environment proved that the algorithmcan not only optimize the clustering function, but also improve the safetyperformance of Ad Hoc network in the non-secure network environment.