Hybridization of Metaheuristics Optimization Algorithm Based Packet Adjustment Rate Model for Congestion Control in MANET

Abstract

Mobile adhoc networks (MANET) includes a collection of independent mobile nodes with dynamic network topology, asymmetry, multi-hop communication, and limited bandwidth. The frequent changes of the network topology and the shared nature of the wireless channel pose significant challenges. Because of restricted resources exist in the network and high data load in MANET; network might be influenced by congestion. This paper presents a new genetic algorithm (GA) with cuckoo search (CS) algorithm, called GA-CS for TCP congestion control in MANET. The GA-CS algorithm makes use of a packet adjustment rate (PAR) mechanism, which control the flow of packets between the cluster heads (CHs) to alleviate congestion. In GA-CS algorithm, the genetic operators are applied for the population initialization of CS algorithm. This integration of genetic operations in CS algorithm helps to achieve proper tradeoff among the exploration and exploitation. The proposed GA-CS algorithm selects the CHs using a weighted clustering algorithm, which selects the CHs based on residual energy, mobility rate, and node degree and transmission rate. The GA-CS algorithm allocates the PAR to the CHs, which enables sharing of data between the clusters to avoid and control congestion. Once the data traffic at the CHs crosses the queue size, the GA-CS algorithm took over its part neutralizing the congestion. A comprehensive simulation takes place to ensure the effective outcome of the GA-CS algorithm and the results are validated interms of throughput, packet loss, delay and congestion control. The attained results ensured the betterment of the GA-CS algorithm over the compared methods.

Keywords: Clustering, Congestion Control, MANET, Genetic algorithm, Cuckoo Search