Research On Tagged-objects Oriented Context Management In RFID Networks

RFID networks have a significant impact on supply chain management for tagged-items to achieve efficient inventory management, anti-counterfeiting, tracking and tracing, etc., through the technology of automatic identification and computer networks. The massive RFID network data, nevertheless, is difficult to utilize efficiently while its volume growing rapidly, i.e. the participants within or outside the supply chain are difficult to communicate with each other due to the absence of specific description for the data chain, or are difficult to use due to the separated distribution of the information chain, or are difficult to reveal values due to the underexposed relationship in the value chain. The main study area of RFID networks limits to the transition management within supply chain recently. Therefore, the study about the value of the RFID data is still inadequate. According to the view of tagged-objects oriented context management in RFID networks, this thesis presents the idea and methods to deal with the problem mentioned above in three aspects: enhancing the descriptive power of the data chain, promoting the efficiency of the information chain, and expanding the scope of the value chain. The contributions of this thesis may have practical significance for effective RFID network and its data value, and may improve the competitiveness of both the supply chain and its participants.This thesis mainly focuses on:1) formally described the data units and supply chain processes based on an abstract model of RFID data unit and SCOR model, respectively, in RFID data chain, to facilitate the communication between the participants or the organization of supply chain in the perspective of life circle of an RFID tag;2) established the ontology for RFID network and context model for tagged-objects to formally describe the commercial state and life cycle transitions and to improve the descriptive power of the chain of RFID data;3) proposed the contextual information integration architecture to publish, discover and collect the contextual information of tagged-objects in the separated distribution in the supply chain, and to integrate the information chain;4) illustrated the index methods for both the current context and the historical context of the tagged-objects to meet the demand of location-based services, approximate keywords error-tolerance, and aggregation events, to efficiently utilize the information chain in RFID network;5) proposed the framework of an information retrieval platform for tagged item based on the contextual information to provide reference information for consumer’s decision-making and to provide a way to show the value of the contextual information;6) provided consumers with location-based or topic-based information retrieval methods for tagged items and their related documents by spatio-approximate-keyword hybrid indices, item language model and latent topic model to expand the RFID network value chain.The innovation and contributions of this thesis includes:First, an context model for RFID-tagged items was proposed to enhance the descriptive power of the RFID data chain. Rather than the device-level contextual information of the RFID system about which has been wildly studied in recent works, this thesis extended the study area of context into information resources in three areas: RFID networks, supply chain processes and Web. Therefore, RFID events, information services and Web documents, etc., are treated as contextual elements in the life cycle transition of an individual RFID tags in the supply chain processes. As a result, the value of the RFID network data are expanded since that the formal descriptive power of context for the business value of the tagged items is improved.Second, a framework and the related methods was described to integrate the contextual information of tagged items in the RFID information chain. An active mechanism for publishing, discovery and collection of public business events, which is defined as the main form to share the RFID context due to the lack of specification of information services discovery for present, is proposed based on ONS and Publish service. Therefore, the efficiency of the RFID information utilization is improved since that the context of RFID tagged items is integrated automatically from the separated distribution of the RFID data chain in supply chain.Finally, an context-base information retrieval platform and relevant algorithms were studied to expand the value for the tagged items in the RFID value chain. By ways of hybrid spatial indices, latent topic model, etc., the value of RFID data were expanded in three aspects:the applying area of information retrieval technology was expanded from Web resources into RFID network resources in supply chain, the study field of context-aware was expanded from device-level situation into business state of tagged items, and the objective of retrieve was expanded from Webpage into bi-directional information which includes both archiving the latent associate documents by provided with tagged items and archiving the tagged items by provided with user-specific keywords.