An Improved Directed Diffusion for Wireless Sensor Networks

Abstract

Directed diffusion (DD) is typical data-centric protocol for wireless sensor networks, all nodes in a directed diffusion- based network are application-aware, and this enables diffusion to achieve energy savings by selecting empirically good paths and by caching and processing data in-network. But at the beginning of routing being established, interest must be flooded throughout the network, it leads heavy traffic. For this reason, an improved directed diffusion (IDD) is proposed in this paper. IDD cluster all nodes into clusters, and select a few nodes as cluster heads the others as cluster members, interest is flooded only among cluster heads, cluster members sense data and transmit their sensing data to its cluster head, cluster head compresses and aggregates sensing data received from its cluster members and sends this sensing data to base station along a reinforcement path. Simulation results show that IDD has higher energy efficient than DD, it can reduce data transmission delay and prolong the lifetime of wireless sensor networks.