Allocation schemes for distributed problem-solving in logic programming

Decomposition and allocation are common problems in distributed systems. The authors present unified decomposition/allocation solutions to cooperative problem solving; these solutions are expressed in logic programming and run on the parallel knowledge-based system. A summary of the parallel knowledge-based system is included. The static allocation schemes described can be realized either automatically or with the advice of a programmer. The schemes maximize the computational resources of a local node and determine the required communication connections of each node within the network. The A* Search program is used to illustrate the allocation schemes. Different decompositions are derived for different schemes.< >