Improved simulated annealing algorithm for task allocation in real-time distributed systems

This paper addresses the problem of task allocation in real-time distributed systems with the goal of maximizing the system reliability, it has been shown to be NP-hard. Many studies have been made to solve this problem without considering the real-time constraint. We first take account of the deadline constraint in order to formulate this problem, and then propose an improved simulated annealing algorithm call adaptive memory-based simulated annealing (AMSA) to solve the problem. The AMSA introduces adaptive factors to reduce the total computation time, and adds memory function to save the recently visited solutions and best solution by now. The effectiveness of AMSA is evaluated by comparing with traditional simulated annealing algorithm. The results show that AMSA can produce “good enough” solution in much less time.