Implementation of K-Means Clustering in Route Discovery in Ad hoc Networks
R Hema, Gayathri Subramanian, M Ajitkumar · 2024
In ad hoc routed networks, a high proportion of RREQ packets cause network congestion, which worsens the data delivery ratio and certain other Quality of Service (QoS) metrics. For instance, in modern ad hoc on-demand distance vector (AODV) protocols control network traffic by advertising detours. It has been established that clustering algorithms are a good substitute for conventional AODV routing. K-means clustering technique is therefore chosen instead of other typical AODV routing equivalents since it is computationally faster than hierarchical grouping with a higher number of variables and produces tighter clusters than hierarchical clustering. Since mobile hosts run on restricted battery resources, the lifespan and performance of the network are increased when energy consumption is reduced. By varying the network size, and the maximum speed of mobile hosts, four performance metrics and energy consumption are assessed for networks with “AODV optimized route discovery using K-means clustering” using the NS-3 simulator. Furthermore, UDP protocol is used at the transport layer and WiFi helper modules enable packet transmission over channels in IP V4-based networks.