Performance Comparison of Proposed Lifetime Maximizing Trees for Data Aggregation in Wireless Sensor Networks

TL;DR

This paper compares two proposed lifetime maximizing tree algorithms, DLMT and CLMT, with existing structures in wireless sensor networks, demonstrating significant improvements in node lifetime, delay reduction, and packet delivery ratio.

Contribution

Introduces decentralized and centralized lifetime maximizing tree algorithms (DLMT and CLMT) that significantly enhance network lifetime and performance over existing methods.

Findings

DLMT extends node lifetime by up to 147%.

DLMT provides up to 13% additional lifetime savings without increasing delay.

Both DLMT and CLMT improve packet delivery ratio and reduce delay compared to E-Span.

Abstract

In this paper a packet level simulator is used to explore the performance of the proposed DLMT and CLMT algorithms under various traffic conditions. Performance of the proposed algorithms is compared with already existing E-Span tree structure. These proposed algorithms tend to extend the node lifetime in order to increase the amount of information gathered by the tree root. Decentralized lifetime maximizing tree (DLMT) features in nodes with higher energy to be chosen as data aggregating parents while Centralized Lifetime Maximizing Tree (CLMT) features with the identification of the bottleneck node to collect data in a central manner among given set of nodes. By choosing Forwarded Diffusion as our underlying routing platform the simulations are carried on J-Sim. Our simulation results have shown that the functional lifetime of event sources can be enhanced by a maximum of 147% when data is aggregated via DLMT and by 139% when data is aggregated via CLMT. Our proposed DLMT algorithm has shown maximum of 13% additional lifetime saving without increasing the delay. Packet delivery ratio has also shown a remarkable increase when the tree depth is considered in these proposed tree structures. Furthermore, the delay is also reduced by using DLMT & CLMT in comparison with E-Span.