[ English | Japanese ]
This dissertation proposes the Information Relocatable Paradigm, effective on a real-space information system such as the Internet of Things or other cyber-physical systems, that handle real-space information generated among distributed locations.
Real space and information space on the Internet are becoming more tightly bound. The major difference between real-space information systems and conventional information systems is the existence of gateway components such as sensors or actuators. The placement in real space of these components, which connect real space and information space, is determined by the target of a sensor or an actuator. Therefore, placement of components in real space affects the placement of the components in the information-space.
Real-space information systems components have four types of roles -- create, store, process and consume -- and their relationships can be categorized into three types: network-based centralized, distributed, and stand-alone. Centralized or distributed systems typically use the client-server model with a design which gathers information to a predetermined component.
The time taken for the communication between components, and the reliability of the communication greatly influences the overall system, if the two components are placed in distant locations, rather than locally. The typical real-space information system is not optimal in terms of performance or reliability because these systems do not pay great attention to the properties caused by the placement of components.
To achieve greater applicability for high resolution real-space information and support diverse real-space information types, we need to develop a high performance real-space information system rather than a simple client-server based system. For this purpose, we need a new communication-optimized architecture that is aware of the placement of the gateway components.
This dissertation pays attention to types of components which act as gateways between real space and information space, and proposes the Information Relocatable Paradigm which places information to a component proximity to a gateway component which has high probability of interest in the information, to reduce the distance between the components which try to store information or use the information. By applying this scheme, the overall system performance can be optimized by relocating information which was previously assumed to be stored at a fixed location.
We choose an RFID information system to apply the new paradigm, revealing the issues in application. Three efforts have been selected to demonstrate our approach to overcoming these issues: establishment of trust relationships between components without using a trusted third party, data security in a RFID user memory, and a repository architecture which relocates information between repositories alongside the item in real space.
By solving issues with these three efforts, we have shown the effectiveness and the feasibility of Information Relocatable Paradigm.
Keio University, Graduate School of Media and Governance
MAUI Project
Ph.D. Dissertation
Back to Index Page
ACADEMIC YEAR
2011 (February 1st, 2012)
NAME
SUZUKI, Shigeya
TITLE
A Relocatable RFID Information System
ABSTRACT
CONTACT
To obtain the dissertation, please contact;
SUZUKI, Shigeya ( shigeya at sfc.wide.ad.jp )