Research For AHM-GAF Algorithm Based On Topology Control Of WSN

The WSN, which integrates four kinds of technologies of the sensor, embedded software, network and wireless corresponds, holds a huge potential utility features in the fields of military and civil applications. It has an increasing significance in the research of computer science. However, due to the unique characteristics of WSN that differ from the traditional network and the current Ad-Hoc network, many new problems have emerged and needed to be solved in the respect of structures of WSN. The features of dynamic variation and limitation of energy that WSN carries have determined the design of topology control being a vital part in its research.The research of topology control contains two aspects of power control and sleep mechanism control. Hereinto, the power control refers to LMA, LMN, CBTC and XTC algorithms etc; the sleep mechanism control concerns with SPAN, CCP, HEED and GAF algorithms etc. According to the requirements of the project, this paper chooses the GAF algorithm to do the topology control in order that the whole network can work with more efficiency and energy-saving.In the study of topology control, GAF (Geographic Adaptive Fidelity) algorithm has been broadly applied. Having analyzed the GAF algorithm, in line with the practical project, this paper has concentrated on and put improvements in the following four aspects:region division,head nodes election,confirmation of head notes and the re-division of virtual region. To the first point, by using hexagons instead of square to divide the network region, it may result in fewer head nodes joined so as to ensure the smooth network connection when under the similar emissive power; to the second point, by taking the surplus energy and the representational character of the collected information from head notes into consideration, it can lead to more surplus energy of the elected head notes and make it much closer to the center of the unit regions. To the third point, by combining the suitable emissive power of head nodes, it indicates that under general circumstances, the length of emissive radius is quite more reasonable between 2 to 3 times of the length of a unit region; then, to the last point, by adapting the way of parallel transfer for the realization of re-division, this will end in a balanced consumption of the whole network whose lifecycle will be nicely prolonged.The simulation outcome results from NS-2 indicated that, compared with GAF, the AHM-GAF algorithm reduced the energy consumption of the head notes in the process of the entire WSN operation and lengthened the lifecycle of the sensor network, meanwhile, increased the accuracy of information gathering and enhances robustness of the entire network. These improved points make the WSN's topology structure more reasonable, and far more applicable to the real network environment.