Neural Network Based Power Control Algorithms for Mobile Ad-hoc Networks

Abstract

This research work presents the ability of simulators in analysis and determination of energy consumption for mobile ad-hoc network. New energy computation methods are developed with neural network. Proposed neural network models are used to overcome the problem of simulator for energy computing due to its complex procedure in measurement of energy consumption values for mobile ad-hoc networks. Mobile ad-hoc network is a special type of wireless ad-hoc network due to its infrastructure and nature of operation. Nodes or devices are moving or changing positions frequently and act as routers in the mobile ad-hoc network. Nodes are communicating with each other through wireless mode without any physical hardware support while moving. Network's topology is changing rapidly and unpredictably and nodes are operated with limited battery in the mobile ad-hoc network. Power control is the necessity for ad-hoc network to run mobile nodes for long periods. Energy optimization is possible by power saving and power control in mobile ad-hoc network. Power saving means to reduce power consumption and power control means to adjust transmission power of mobile nodes to minimize energy consumption in mobile ad-hoc network. Power control reduces data retransmission probability with a good assignment of transmission power and node guarantees it transmission in a low number of attempts. Main focus on the energy consumption not only because that it is the key issue in the mobile ad-hoc network, but also, it is observed from practical experiments that energy consumption problem also affects network performance metrics for the mobile ad-hoc network. Power is consumed even if the frame is not used by the nodes because it was intended for another destination in network. Nodes have consumed energy even if intermediate nodes are not in the communication radio range of each other and need to rely on multi hop transmissions in network. Transmission with optimum transmission energy of nodes is used to reduce the energy consumption of nodes in mobile ad-hoc network. Transmission energy is inversely proportional to remaining energy of nodes in the network. Remaining energy is the energy of the nodes after transreceiving data in mobile ad-hoc network. Simulators are used for analysis of energy consumption among nodes in the network. Simulator is more time consuming and complex procedure to test data for any unknown value, once program run in the simulation environment. Neural network model is a mathematical model to determine the node energy consumption for any unknown value in an ad-hoc network. This model is alternative solution of simulator for energy computation in ad-hoc network. A neural network concept is approaching to estimate the energy consumption and reduce the complexity in computation of power consumption with simulator for mobile ad-hoc network.