Research on an Adaptive Algorithm of Service Selection in Pervasive Computing

An adaptive genetic algorithm of service selection in pervasive computing is presented in this paper. By means of matrix encoding, this algorithm carries out selection, crossover and mutation operations of genetic algorithm, with matrix as individual chromosome and matrix array as gene. Based on the elitism selection strategy and adaptive strategy, this algorithm replicates the optimal individual of every generation to the next generation, and adaptively adjusts crossover and mutation probabilities according to the fitness value of individual, which is aimed at protecting the individuals with high fitness and weeding out those with low fitness. Simulated comparison proves that this algorithm is of good convergence, which also shows more stable optimization ability, compared with ordinary genetic algorithm and random selection algorithm.