Method of Power Adaptation for Signals Emitted in a Wireless Network in Terms of Neuro-Fuzzy System

Attenuation of the signal propagating between the emitting and receiving antennas in a wireless network requires adaptation of the power of the emitted signal for the stable functioning of the network. The random character of the signal attenuation, frequently leading to the so-called fading of the signal is a challenge demanding prediction of the signal intensity in each next time period. We argue that the problem can be solved employing the fuzzy logic by loading the average values of signal attenuation during the n previous time intervals into the input of the neuro-fuzzy system. A new method for adaptation of the power of radio transmitting devices based on the application of neuro-fuzzy systems is described. The synthesized neuro-fuzzy system has the capability to select the set of input parameters depending on the features of the wireless network. Training of such a system (for four input values) provides the prediction of the signal intensity with the accuracy of 97–98%. The proposed method provides the adaption of the signal power in accordance with the requirements for the bandwidth capability of the radio path.