Collection of wireless nodes like mobile, laptops, palmtops etc. can creates a dynamic wireless network without using any existing infrastructure known as Mobile Ad-Hoc Network (MANET). MANET performs peer to peer interactions among the wireless nodes participating in the network. Each node in the network must be able to function as a router as well to relay the packets generated by other nodes. Due to the mobility of nodes network topology changes frequently and the topology changes information must update every time to all the nodes in the network. Limited bandwidth and transmission characteristic impose administrative and control information to update the dynamic charge in the network topology, due to the mobility of the wireless nodes. Information update at some fixed time interval can cause unnecessary traffic in the wireless network so to make adaptive the time interval of these messages is a technique to improve the performance of the network. In this paper a soft computing technique Artificial Neural Network based reactive AODV routing protocol is proposed to determine the frequency of hello interval to improve the performance of the network.