Research On Ad Hoc Network Clustering Algorithms Based On Signal Strength

Ad Hoc network has no center, self-organizing, jump routing characteristics etc, these make it very suitable for some special geographical environment, such as earthquake, typhoon. It is the characteristic that the Ad Hoc network group-dividing becomes unique. The first problem of group-dividing is to solve how many amounts is the best. Excessive number of clustering makes system cost and data transfer delay too much. Too little number will cause excessive busy of head, excessive waiting of information. Good clustering algorithm shall make the Ad Hoc network have the moderate number, reasonable structure. Therefore, a suitable algorithm will be particularly important for Ad Hoc network.Currently, many experiences have been accumulated for Ad Hoc networks clustering algorithm research. From different point of view, it will get different group-dividing method. The Lowest-ID algorithm maintains the network group-dividing fast. The Highest-Degree algorithm minimizes the number of clustering, The Lowest-speed algorithm maintains group-dividing stability. Adaptive on-demand weighted algorithm (AOW) integrated approach of various clustering algorithms, giving reasonable weight for each method according to its importance. Basing on this thoughts, this paper presents a new algorithm signal strength adaptive On-demand Weighting algorithm(SAOW). The algorithm applied the signal strength in the clustering process, and regarding the sum of the weights of signal strength Sn, and the the stability of Cn as the two main important factors during the election of cluster header.In this algorithm, each node have signal strength weights, information weights, the battery energy weights and weighted of stability, meanwhile the group structure is improved using signal intensity. It can not be well distinguished relative motion from absolute motion in the AOW algorithm. This paper aims to resolve poor performance on mobility, the computational complexity of problems in the AOW. The choice of head has a larger change, but it has a closely relation between the changed factors and the original factors. After then, the four stages of the Ad Hocnetworks division, and the problems encountered at all stages are discussed. Finally, the performance test for SAOW algorithm is given using the simulation environment of NS2. It shows that SAOW has a better stability of AD hoc network comparing with AOW algorithm.