Evaluation of heuristic path selection algorithms for multi-constrained QoS routing

Multi-constrained quality of service (QoS) routing aims to find a path in a network that satisfies multiple path constraints. It is known to be NP-complete and many heuristic algorithms have been presented to solve it sub-optimally. Recent promising proposals include the TAMCRA and limited path Dijkstra´s algorithm (LPDA) which are both based on extended Dijkstra´s algorithms using non-linear path functions, limited path heuristic algorithm (LPHA) which is an extended Bellman Ford algorithm, and improved limited path heuristic algorithm (ILPHA) which improves LPHA using non-linear path functions for path selection. In this paper, we give a simulation study for comparisons of these heuristic path selection algorithms. It is shown that path selection strategy in ILPHA can significantly improve the performance of LPHA and the heuristic algorithms with dominating area non-linear path function (ILPHA-DomArea, LPDA-DomArea) outperform the heuristic algorithms with the max weight based nonlinear path function (ILPHA-MaxXY, TAMCRA).