Nearly Monotonic Problems: A Key to Effective FA/C Distributed Sensor Interpretation?

Abstract

The functionally-accurate, cooperative (FA/C) distributed problem-solving paradigm is one approach for organizing distributed problem solving among homogeneous, cooperating agents. FA/C agents produce tentative, partial solutions based on only local information and then exchange these results, exploiting the constraints that exist among their local subproblems to resolve uncertainties and global inconsistencies. A key assumption of the FA/C model has been that the local solutions can substitute for the raw data in determining the global solutions. This is not the case in general, however. Does this mean that researchers' intuitions have been wrong and/or that FA/C problem solving is not likely to be effective? We postulate that some domains have a characteristic that can account for the success of exchanging mainly local solutions. We call such problems nearly monotonic. The basic idea is that while belief and solution membership are nonmonotonic with increasing evidence, partial solut...