A NOVEL CLUSTER-BASED ROUTING MODEL FOR EFFICIENT INTERNET OF THINGS COMMUNICATION

The Internet of Things network is a structure of interconnected gadgets that connects with each other and recognize information about the environment. The integration, network and transmission rules of this network device are highly dependent on the routing protocol used in the architecture. The Internet of Things has become a necessary principle in human life, and routing protocols that aid communication are necessary to realize this vision. This research is an extension to further increase the rate of packet delivery to destination. Earlier research enabled the use of cluster-based routing protocol for energy efficiency forgoing the relevance to traffic pattern which is the sole essence of Internet of Things networking. This research hampers on the packet delivery rate, increase on packets delay, reduction in data collision as well as reduced communication overhead. This research made use of the Anaconda 3 (Python Distribution) as the programming language, MatLab 2023 for the simulation, K-Means algorithm for clustering and designation of master nodes for clusters and the Ad-hoc On-demand Distance Vector (AODV) algorithm for route discovery, packet forwarding and routes maintenance. The idea of this algorithm is to improve the performance of Internet of Things network in terms of packet transmission. The computational time and accuracy level results were calculated using contrasting numbers of sensor nodes and K-values in checking the percent error of each simulation, the accuracy level of clustering was 0.88% and the accuracy of the route discovery was 0.91%. To check the scalability of the network, a comparative analysis was demonstrated with Matlab simulator using various parameters: area size 200m, number of nodes 1000, transmission range 1000m, packet size 5, which gave 87.2% performance, the packet delivery ratio showed an accuracy of 0.98% of high and 0.78% of low, the packet delay ratio showed 0.18% of high and 0.04% of low and the total performance of the system showed an average of 98.4% accuracy in performance