Rule-based information distribution systems: near-optimal rule sets generation

Information distribution rules (IDRs) can be utilized to effectively and efficiently control event-driven information flow between objects in distributed-object systems. The problem with IDRs is that as the configuration of the distributed system gets larger (i.e., large number of objects, large number of possible events, more need for information, large number of communication constraints, etc.), it is very difficult to generate and debug IDRs manually. Additionally, having faulty IDRs can lead to serious undesirable situations. This paper presents an algorithm to generate near-optimal IDRs for efficient and effective information flow. Illustrative examples as well as complexity analysis of the behaviour of this algorithm are included in the paper.