Managing Collaborative Protocol for Wireless Sensor Networks

Managing Collaborative Protocol for Wireless Sensor Networks

Thasayu Soisoonthorn and Sirichai Rujipattanapong

Department of Electrical Engineering, Faculty of Engineering King Mongkut University of Technology North Bangkok, Bangkok 10800 Thailand Tel: +66-2-9132500 ext 8519, Fax: +66-2-5857350 E-mail: srujipat@kmutnb.ac.th

Abstract— Wireless sensor network protocols are mostly designed with each node independently processing and transmitting small amount of data to base station. Thus, they may be unsuitable for some applications, such as target tracking. In this kind of application, it requires data from many nodes to track target accurately. Hence, transmitting such large data directly to base station is not efficient. It can cause contention, collision, and reduced performance of the whole network. Thus local collaboration among nearby nodes is necessary. In this paper, the protocol that helps manage the collaboration among neighbor nodes is proposed. This protocol is named managing collaborative protocol (MCP). Based on the modification of TTDD, its design goal is to reduce the overhead of cluster formation. Compared to both DCTC and BACKOFF, using ns2, MCP reduces the number of transmissions; delivery success rate is higher, and area coverage of the event is better.

I.

INTRODUCTION

Manufacturing technology of electronic devices improves at a tremendous rate. As a result, sensor nodes are more capable and consume less energy. However, sensor node still cannot harvest energy directly from environment. Therefore, energy consumption must be taken in to account when designing protocol for sensor nodes. Thus, it is preferable to design protocol that can manage collaboration and balance energy utilization among sensor nodes to achieve the objective of the application. Related works are summarized here. In UW-API [1] and Dynamic Clustering [2] both use collaboration. However, they both have shortcomings. That is, each node has specific role: cluster head...