A heuristic genetic algorithm for the single source shortest path problem

This paper addresses one of the potential graph-based problems that arises when an optimal shortest path solution, or near optimal solution is acceptable, namely the Single Source Shortest Path (SSP) problem. To this end, a novel Heuristic Genetic Algorithm (HGA) to solve the SSSP problem is developed and evaluated. The proposed algorithm employs knowledge from deterministic techniques and the genetic mechanism to achieve high performance and allow consistent convergence. In addition, the proposed HGA is implemented and evaluated using a developed software tool that is easily amenable for future extensions and variations of our HGA. The schema introduced in this proposal depends on starting with initial population of candidate solution paths constraints as an alternative of a randomly generated one. To preserve the high performance candidate solutions, the HGA also uses a new heuristic order crossover (HOC) operator and mutation (HSM) operator to keep the search limited to feasible search domain. Simulation results indicate that the developed HGA is highly efficient in finding an optimal also quantify the effect initial population size and the increase of generation numbers.