Frequency Domain Equalization using EHW based Neural Network

This paper describes an evolvable hardware (EHW) system for modelling a adaptive neural network. We have developed an adaptive RBF neural network which can be used to model plenty of systems. This approach is particularly suited to applications requiring ability to cope with time-varying problems and realtime timing constraints. FPGA as a reconfigurable hardware is used for hardware implementation purpose of EHW_based NN. The GA is used to determine the characteristic of neurons in RBF and the adaptation of weights is done with a steepest descend method. Simulation results for an adaptive frequency domain equalizer in digital mobile communication are given. Our system shows better results comparing to time domain one.

کلیدواژه ها:

channel equalization ، frequencydomain equalization ، evolvable hardware ، adaptive NN ، Genetic Algorithm