IOSR Journal of Engineering
Wireless adhoc network support rapid on demand and adaptive communication among the nodes due to their self configurable and autonomous nature and lack of fixed infrastructure. Since adhoc network rely on the collaboration principle the issue of key distribution and effective group by management in such networks represents two of the most important problem. This paper explore in depth the requirements of this set of applications scenarios and purposes from machine learning based on demand routing paradign. Its primary benefits are flexibility to optimize routing over a variety of property such as route length easy recovery after node failure and native support for sink mobility. As efficient approach to it's consider routing algorithm in which network connectivity is determined in the process of establishing routes.