Task allocation algorithms for maximizing reliability of heterogeneous distributed computing systems

The rapid progress of microprocessor and communication technologies has made the distributed computing system economically attractive for many computer applications. One of the first problems encountered in the operation of a distributed system is the problem of allocating the tasks among the processing nodes. The task allocation problem is known to be computationally intractable for large task sets. In this paper, we consider the task allocation problem with the goal of maximizing reliability of heterogeneous distributed systems. After presenting a quantitative task allocation model, we present a least-cost branch-and-bound algorithm to find optimal task allocations. We also present two heuristic algorithms to obtain suboptimal allocations for realistic size large problems in a reasonable amount of computational time. Simulation was used to study the performance of the proposed algorithms for a large number of problems. Also, performance of the proposed algorithms has been compared with a well-known heuristics available in the literature.