Approach for Multiple Constraints Based Qos Routing Problem of Network

The Intelligent evolutionary algorithm-probability learning based algorithm is applied to multiple constraints Qos(Quality of Service) routing problem of network. The objective function is to minimize the cost and meet multiple constraints. The procedure and the choice probabilitypsilas update strategy of PBIL algorithm are designed according to the characteristics of multiple constraints Qos routing problem. During the evolutionary process of the algorithm the constraint equation is dynamic adjusted to avoid unnecessary search and save the search time. Each invalid path is recorded as a constraint to guide the next iteration. And the probability is updated according to the solutionspsila fitness. The node linkage which belongs to the excellent solution has greater probability to be chosen. The modified PBIL algorithm is tested on a network topology and some routing demands is considered. The good adaptability, validity and stability performance are fully shown by the results.