Joint channel and data estimation: genetic algorithm based blind equalization
Frank Michael Caimi, D. Wang · 1999
A genetic algorithm (GA) based blind equalization method is presented for joint channel and data estimation. The GA is an optimization technique using natural selection and evolutionary processes that searches for solutions of the problem through phases of evaluation, reproduction, crossover, and mutation repeatedly. The most important advantages of these algorithms are parallel search capability, convergence to global optimum, and reduced problem-dependence. Different schemes for each of above phases have been considered in achieving better results. Computer simulations have shown that for all the assumed situations, the presented algorithms establish the channel model and decode the transmitted data with satisfactory precision. Additional improvements for the GA approach are also presented, including GA coupled with a gradient based algorithm, neural network based calculation for the improvement of computational efficiency, and the genetic programming based modeling for nonlinear channels.