A parallel Tabu search algorithm for the 0-1 multidimensional knapsack problem

The 0-1 multidimensional knapsack problem (0-1 MKP) is an NP-complete problem. Its resolution for large scale instances requires a prohibitive processing time. In this paper we propose a new parallel meta-heuristic algorithm based on the Tabu search (TS) for the resolution of the 0-1 MKP. We show that, in addition to reducing execution time, parallel processing can perform automatically and dynamically the settings of some TS parameters. This last point is realized by analyzing the information given by cooperative parallel search processes, and by modifying the search processes during the execution. This approach introduces a new level where balancing between intensification and diversification can be realized.