Intelligent Framework for Qos Optimization in Manet Using Soft Computing Models

Mobile Adhoc Networks (MANET) is a heterogeneous self configuring networks that changes topology often. Due to mobility nature of wireless nodes, routing information needs to be frequently updated. As nodes are wireless in nature, there exists enormous security threats that disturb the deployment and maintenance of MANET. To alleviate such problems, secure agent based multicast routing protocol for wireless network is proposed which is an extension of existing AQMRA. Group communications require stringent QoS parameters and they have to be kept in rigorous bound. In the proposed work Secure ANFIS based QoS Multicast Routing Agent (Secure AQMRA), QoS input parameters namely node speed, link delay, residual power, packet loss, bandwidth, link utilization are optimized to achieve better scalability. Secure agents are configured in each and every node in the network. This protocol guarantees the security, scalability and high reliability of an established route. The proposed Secure AQMRA protocol performs better than the existing AQMRA protocol in terms of end to end delay, control overhead and Packet Delivery Ratio. KEYWORDS-MANET, Neighbor discovery, Multi-agent, Secure Routing, Neuro Fuzzy, Optimization, QoS.