Issue No. 03 - March (1991 vol. 40)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/12.76410
<p>A graph-based solution to the mapping problem using the simulated annealing optimization heuristic is developed. An automated two-phase mapping strategy is formulated: process annealing assigns parallel processes to processing nodes, and connection annealing schedules traffic connections on network data links so that interprocess communication conflicts are minimized. To evaluate the quality of generated mappings. cost functions suitable for simulated annealing that accurately quantify communications overhead are derived. Communication efficiency is formulated to measure the quality of assignments when the optimal mapping is unknown. The mapping scheme is implemented using the hypercube as a host architecture, and results for several image graphs are presented.</p>
processor assignment; scheduling; communication efficiency; hypercube architecture; link assignment; multicomputers; mapping problem; simulated annealing optimization heuristic; automated two-phase mapping; process annealing; parallel processes; processing nodes; connection annealing; traffic connections; network data links; communication conflicts; cost functions; communications overhead; image graphs; graph theory; hypercube networks; simulated annealing.
S. Bollinger and S. Midkiff, "Heuristic Technique for Processor and Link Assignment in Multicomputers," in IEEE Transactions on Computers, vol. 40, no. , pp. 325-333, 1991.