Adaptive mobile GIS and applications

Abstract

The increasing need for using maps to communicate and display geospatial information for location-based services and mobile applications has brought many challenging issues. Because of limited network speed, bandwidth, lack of efficient data representation structures, and incompatibility of data formats, the distribution of a rich set of geospatial information over the Internet and wireless environments is, at present, somewhat limited. The limitations include difficulty in handling large amounts of spatial data, high processing cost, and weak data protection measures. Although some methods have been used to relax the above constraints, an example being the representation of spatial data by static images with different pre-set scales, these methods are mostly vendor proprietary and not adaptive to different applications and user's requirements for mobile computing in heterogeneous networking environments. In the context of the mobile Internet and wireless devices, the openness, extensibility and flexibility of spatial data models used in applications are important in terms of maps on demand, data storage, data transfer and data processing. Hence the provision of an open, extendable and flexible spatial data model should, in all likelihood, substantially enhance the performance of mobile device spatial data transmission and display. The research presented in this thesis aims to develop an adaptive mobile Geographical Information Systems (GIS) framework, designed to address the following important issues: (a) portability and sharing of geospatial data; (b) intelligent processing of geospatial data; and (c) performance of geospatial data transmission. Solutions are sought by developing mobile spatial data modelling techniques with free, standard and non-proprietary formats and proposing methods and algorithms for delivering and displaying mapping information adaptively, in accordance with users' various and varying requirements. The adaptive mobile GIS framework is based on a three-tier client/server architecture and consists of a client application tier, middleware tier and data management tier. The client tier mainly handles effective and efficient presentation of spatial information to the users. The middleware tier supports adaptation in terms of personalization, security, and all geospatia] information tasks such as map content request and complex spatial query. The data management tier is responsible for storing and accessing spatial information, using an open source standard relational database management system (RDBMS), which implements spatial extensions and follows the specification of the Open Geospatial Consortium (OGC). The prime objective of this research is to derive an optimized mobile spatial data model with high adaptability as regards the representation and processing of georeferenced map images and associated attributes for adaptive mobile GIS frameworks. It is intended to be an efficient mechanism for provisioning mobile GIS applications. An efficient spatial data transferring algorithm for delivering spatial data, named by the author, Adaptive Base Block of Object Transmission (ABBOT), has been developed and refined for various existing connecting environments. A highly efficient indexing scheme, Geo-Planar Indexing (GPI), has been derived for fast retrieval of geospatial data under the standard RDBMS environment without using expensive R-Tree spatial access methods in mobile working environments. Extensive experiments have been conducted to test the efficiency and usability of the three component tiers of the adaptive mobile' Gis framework. Preliminary results have been achieved. One important technique used in the client application tier is the employment of Scalable Vector Graphics (SVG) to present geospatial and non-graphic data on mobile devices. A technique used in the data management tier is the utilization of Geography Markup Language (GML) to encode spatial data. A mobile test site was set up to demonstrate the use of Pocket IE (a web browser developed by Microsoft) to view SVG-based spatial data in a Pocket PC environment. A proof-of-concept testing system was set up to demonstrate the use of GML and SVG in the process of online editing spatial data. A software package, named by the author, libMobileGIS, has been successfully compiled during the system prototyping process. The package is a set of dynamic link libraries, comprising 6 open source programs, the aims of which are to target mobile devices using Microsoft Windows Mobile OS. This package is the first of its kind to be made available to the public, by the mobile GIS community and forms the foundation for serious GIS mobile device development. The successful testing of a series of complex mobile device spatial analysis queries concludes that the proposed adaptive mobile GIS framework is viable for designing and developing a fully functional mobile GIS with good performance in spatial data handling and complex spatial query operation.