DocumentCode :
3025046
Title :
Task assignment by parallel simulated annealing
Author :
Witte, E.E. ; Chamberlain, R.D. ; Franklin, M.A.
Author_Institution :
Comput. & Commun. Res. Center, Washington Univ., St. Louis, MO, USA
fYear :
1990
fDate :
17-19 Sep 1990
Firstpage :
74
Lastpage :
77
Abstract :
Simulated annealing for obtaining approximate solutions to combinatorial optimization problems is addressed. The serial algorithm, however, can require extensive computation time. Most parallel algorithms for simulated annealing are problem-specific and/or violate the serial decision sequence, thereby allowing errors not present in the serial algorithm. Maintaining the serial sequence is necessary to prove that the algorithm converges to a global optimum solution when allowed to reach equilibrium at each temperature. A parallel algorithm which is both problem-independent and maintains the serial decision sequence is presented. The parallel algorithm uses the concurrency techniques of speculative computation to achieve speedup which can exceed log2 P, on P processors. For three problems investigated, the average speedup on eight processors was 2.6
Keywords :
parallel algorithms; simulated annealing; approximate solutions; combinatorial optimization problems; concurrency; parallel algorithms; parallel simulated annealing; problem-independent; serial decision sequence; speculative computation; task assignment; Computational modeling; Computer simulation; Concurrent computing; Cost function; Data structures; Parallel algorithms; Parallel processing; Simulated annealing; Temperature control; Temperature distribution;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Design: VLSI in Computers and Processors, 1990. ICCD '90. Proceedings, 1990 IEEE International Conference on
Conference_Location :
Cambridge, MA
Print_ISBN :
0-8186-2079-X
Type :
conf
DOI :
10.1109/ICCD.1990.130165
Filename :
130165
Link To Document :
بازگشت