AN INNOVATIVE LBS DATA SERVICE INFRASTRUCTURE
Location based services (LBS) is one of the quickest developing technology in the past 20 years. But its QoS (quality of service) is one of critical issues and bottleneck handicapping further development and more application. The current LBS QoS problems include: data especially non-spatial data is out-of-data and/or error, data updating completely by special data provider cannot meet the increasing user demand; what the consumers paid for is not exactly what they wanted most; LBS service and system design are not ego centric but allocentric; and no valid alternative service or measures when network connection is unstable. This paper presents an innovative cell-based multi-level hierarchy data-service infrastructure which is composed of ubiquitous computing technique, GIS and cognition science can partly solve aforementioned problems. In the data-service infrastructure spatial and non-spatial are stored and indexed in cells in various granularity according to various granularity; data provider and individual provide heterogeneous spatial and non-spatial data through web services; data servers provide mobile client-end not allocentric but ego centric service based on context; there is some data replication being stored in mobile client-end, and when the client-end submits query to servers, the servers will check data version of ROI (Region of Interest) in database with the one stored in client-end. The mobile client-end will then decide whether download data package or not; the query result will be packed in one or more package(s) and the package(s) will be transferred sequentially according to their importance grade; the data flow between servers and mobile client-ends is bidirectional; data service consumers are data verifier, data evaluator and data provider too. In the data service infrastructure, lower threshold makes every netizen become data provider and contribute to QoS improving; celled-based data storage makes data transfer and data replication easy to implement; data sequential transferring according to importance grade make users(especially emergence rescue) can get essential data service in the first time in unstable network connection condition, and data evaluation based on spatio-temporal data mining provide an efficient and objective information extraction method. Experiment proves that our data-service infrastructure is efficient and effective in navigation, Wayfinding and emergence rescue.
Mobile GIS Data Infrastructure Computer Vision Data Integration Location Based Services
Danhuai GUO Weihong CUI
Institute of Remote Sensing Applications, Chinese Academy of Sciences P.O.Box9718, Beijing 100101, China
国际会议
北京
英文
4202-4207
2008-07-03（万方平台首次上网日期，不代表论文的发表时间）