ADAPTIVE GRID MULTIHOP ROUTING PROTOCOL (AGMRP) FOR A HOMOGENEOUS WSN USING SPECTRAL GRAPH PARTITIONING TECHNIQUE

Abstract

Energy limitation is a major constraint affecting the performance of a WSN due to the limited
energy of the nodes. The sensor nodes are mostly battery-powered and are deployed in remote
or hostile surroundings, whereby it is difficult to recharge or replace the batteries regularly.
Thus, hindering their ability to carry out their functions efficiently. As such, reducing energy
consumption and prolonging the lifetime of the nodes becomes necessary. In this paper, an
energy-efficient routing algorithm was developed utilizing the spectral graph-partitioning
technique based on Normalized cut (Ncut) that optimally partitioned the network into clusters
(grids). An unequal clustering mechanism for mitigating the hotspot problem was
incorporated into the routing algorithm. The algorithm was implemented in MATLAB
R2018a and was evaluated using network lifetime, energy consumption, and some alive nodes
as performance metrics. Simulations were carried out in two scenarios, and results showed
that AGMRP outperformed other algorithms.