Publication Date

12-2012

Type of Culminating Activity

Thesis

Degree Title

Master of Science in Computer Science

Department

Computer Science

Major Advisor

Sirisha Medidi, Ph.D.

Abstract

Wireless sensor networks (WSNs) are event-based systems that rely on the collective effort of several sensor nodes. When all nodes in an area sense an event and transmit that data, it causes sudden traffic bursts, which are spatially-correlated and lead to network congestion. Congestion can cause an increase in the amount of data loss, energy consumption, delay data transmission, and hinder network performance. To improve performance of event-driven applications, there arises a need for protocols that can reduce congestion and energy consumption. Existing protocols for sensing multiple events either handle congestion control or spatially-correlated contention, but not both, which can degrade network performance in terms of packet delivery ratio, latency, and energy consumption. Motivated primarily by the challenge to improve the performance of event-driven applications, we propose an energy efficient protocol to mitigate congestion that improves data delivery and reduces latency. This protocol mitigates congestion by dispersing network traffic using a forwarder selection mechanism that forces event reports from different nodes to disperse along different paths to the base station. Our protocol also reduces spatially-related contention by partitioning the sensors into different groups. All the sensors in a particular group cover the region of interest together, and these groups are scheduled in such that only one group is active to transmit the data at any given time. We implemented our protocol using the NS2 simulator for evaluating its performance. Results show that our protocol has significant improvement in the packet-delivery ratio, latency, and energy savings.

Download

Share

COinS