Machine Learning Algorithmic Approach for Handling Link Asymmetry in Wireless Mesh Access Network of Mobile Ad Hoc Network

Mobile Ad hoc Network (MANET) is designed for dynamic networking at less expensive network establishment for various sudden needs of wireless mobile communication. The constituent nodes of MANET are deployed for different purposes of mobile communication. So these nodes are heterogeneous in its configurations and functions like transmission range, node life time, signal strength and etc. Depends upon the internet service requirements MANET is connected to Internet. The internet service provider's access point nodes and few nodes of MANET which are all connected to the internet service provider's access point nodes comprise Wireless Mesh Access network (WMAN). The internet service provider's access point nodes are also heterogeneous in its configurations and functions. So the WMAN of MANET is entirely heterogeneous. The transmission capabilities of the node decide the reliable connectivity of it to the network. Promising such reliable connectivity in heterogenetic transmission range network is a challenging design of network protocol. The proposed system confirms robust connectivity among every node in such network by powering them with artificial intelligence in the process of networking to the hidden neighbour nodes. This artificial intellectual design is achieved by applying machine learning algorithm. Allowing the communications between the devices from different layers of network provides necessary data to apply such machine learning algorithm and to bring the strong connectivity among nodes. The mathematical analysis on this algorithmic approach produces quantitatively better results in the network performance measuring parameters in comparison with previous research works carried out in the aforementioned challenging design of network protocol for WMAN of MANET.