Novel adaptive DCOPA using dynamic weighting for vector of performances indicators optimization of IoT networks

Abstract

Clustering in the Internet of Things (IoT) involves organizing devices into groups to streamline network management and optimize resource utilization, including Internet connections, energy usage, coverage, quality of service, and connectivity. DCOPA (A Distributed Clustering Based on Objects Performances Aggregation for Hierarchical Communications in IoT Applications) is a recent distributed clustering protocol based on a timer for cluster formation where the election of Cluster Heads (CHs) is modeled as a multicriteria problem. In this paper, three contributions are presented. Firstly, the DCOPA protocol is analyzed with a focus on its multi-criteria aggregation function T(i) which directly contributes to the election of the CHs and the formation of the network’s clusters. This is then followed by an in-depth analysis of the impact and the variation of the weights assigned to the two aggregated criteria which are the energy and the distance from the base station. A verification of the scalability, load balancing and distribution of the clusters and CHs will follow. Secondly, a new formal notation for the performance analysis, specifically focusing on the mortality and lifetime based on the Vector of Performance Indicators (VPI), will be introduced for IoT. As a third contribution, a revised version of DCOPA is introduced called ADCOPA (Adaptive DCOPA Using Dynamic Weighting for Vector of Performances Indicators Optimization of IoT Networks). ADCOPA is based on a new property which is the dynamicity or the variability of the weights of the criteria used in the election function of CHs. The simulation results show that the ADCOPA algorithm, which dynamically adjusts the weights of the criteria during the network’s lifetime, outperforms the DCOPA algorithm. The latter uses static weights for the criteria that remain unchanged for the entire lifetime of the network. This confirms that the ability to dynamically adjust the weighting of the criteria is an important factor in achieving better performance.