Modified GA Based Routing Protocol for Efficient QoS Routing for MANETs

With the recognition of ad-hoc networks, several routing protocols are designed for route discovery and also the route maintenance. Because mobile ad-hoc networks differs from the traditional wired networks. They have certain unique characteristics which can cause difficulties for providing QoS in such networks. The unique characteristics are dynamically varying network topology, lack of precise state information, shared radio channel, limited resources accessibility, hidden terminal problem and insecure medium. Design of efficient routing protocol that satisfies the one QoS metric is difficult. Generally used QoS metrics are end to end delay, throughput, packet delivery ratio, and jitter. Designing of multi constrained QoS routing protocol is still difficult. Because routing protocol has to satisfies the many QoS metrics at a time. Existing genetic algorithms based routing protocols can give solution for multi constrained problem. But it is not giving efficient results. This paper proposes a modified genetic algorithm based routing protocol, and performance is compared with the existing algorithm. Simulations are performed in Network Simulator (NS-2).Simulation results shows that modified and proposed algorithm is giving optimized results than existing one for the specified scenario. While doing the simulation TCP,CBR and video sources are connected seperatly and the performance of two algorithms are compared with delay, throughput and packet delivery ration metrics TCP, UDP and video sources and connected separately and simulation is performed. Simulation results show that modified and proposed genetic algorithm is giving efficient results than existing genetic algorithm based routing algorithm.