Neuro-fuzzy control of data sending in a mobile ad hoc network

Abstract

The article is aimed at data transmission efficiency increase in a mobile ad hoc network, functioning to ensure the exchange of information in emergency situations. The algorithm for neuro-fuzzy control of data sending intensity in this network is proposed. The algorithm provides the measurement of latency period confirmation and the calculation of a moving average of this value, as well as a periodic recalculation of an inter-packet gap by performing neuro-fuzzy inference procedure. By using the algorithm an adaptive change of data sending intensity is provided in the framework of the established logical connection. Simulation experiments showed that the use of the proposed algorithm reduces the number of retransmissions and the average transmission time of data streams in a test network.

Keywords: mobile ad hoc network, emergencies, the intensity of data sending, algorithm