Adaptive Convergecast by Distributed Topology Switching

Suchetana ChakrabortySushanta Karmakar

Convergecast in a wireless sensor network is a process in which eachsensor node senses the environment and forwards that information to a base stationin some way. For correct data gathering using convergecast there shouldbe no data loss and no delivery of redundant data. Sensors can form a spanningtree rooted at the sink (base station) to perform the convergecast in an efficientway. Leaves of the tree can sense and forward data independently. Howeveran internal node forwards data to its parent only after receiving data from allits children. It has been observed that a Breadth-First-Search (BFS) tree is abetter choice for convergecast under low system load because the depth of anynode from the root is always minimum. However under higher load a Depth-First-Search (DFS) tree may be a better option as the degree of any node in aDFS tree is generally lower than that in a BFS tree. Hence load of each node islower in case of a DFS tree than that in a BFS tree. Therefore it may be desirableto dynamically switch between a BFS tree and a DFS tree based on load.In this paper we propose a scheme for adaptive convergecast that dynamicallyswitches between a BFS tree and a DFS tree. The switching mechanism remainstransparent to the convergecast. Also each convergecast message is correctlydelivered to the base station eventually without any loss or redundancy.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: