Self-Organized P2P Approach to Manufacturing Service Discovery for Cross-Enterprise Collaboration

Abstract

The combination of service-oriented architecture (SOA) and peer-to-peer (P2P) architecture plays a promising role in distributed manufacturing environments in that the peer service can be used to facilitate the integration and discovery of distributed manufacturing resources and achieve communication and collaboration across distributed virtual enterprises. However, the large size, dynamic nature, and heterogeneous expression of distributed manufacturing resources bring forth a serious challenge in scalability and efficiency. This paper presents a self-organized P2P framework that supports scalable and efficient manufacturing service (MS) discovery for cross-enterprise collaboration by forming and maintaining autonomous enterprise peer groups (PG). Each enterprise exhibits as a peer that provides some sharable MSs that are represented comprehensively and formally with a generalized ontology. Each enterprise PG dynamically clusters a set of enterprise peers offering semantically similar MSs, and elects the most reputed peer through multicriteria trust evaluation as its core (i.e., super peer, SP). Then, a MS request can be first routed to the suitable SP and further to its leaf peer in a systematic way, thus supporting efficient service discovery. A prototype system is implemented on JXTA for real application and validated through an experimental case study.