Blocking probability is one of the key factors to evaluate the routing and wavelength algorithms for intelligent optical network. Two kinds of dynamick-shortestpath (DkSP) algorithms were designed. One is based on L...
详细信息
ISBN:
(纸本)081946449X
Blocking probability is one of the key factors to evaluate the routing and wavelength algorithms for intelligent optical network. Two kinds of dynamick-shortestpath (DkSP) algorithms were designed. One is based on Linear Link Weight Function (LW) and the other is based on Piecewise Linear Link Weight Function (PLW). It was found that the two kinds of DkSP can significantly decrease the blocking probability of optical network comparing to the static kSP for the same number of alternate routes. Compared to routing with LW, the coefficient of PLW has larger effect on the blocking probability of optical network when the number of alternate route is small, but the effect is weakened with the increase of the number of alternate route. As far as the two kinds of DkSP algorithms are concerned, DkSP with PLW has some advantage over DkSP with LW on decreasing the blocking probability. It was also found that the optimized performance can almost be got by DkSP with only 2-4 alternate routes for NSFNET.
暂无评论