Adaptive, reliable, and accurate positioning model for location-based services

Abstract

This thesis presents a new strategy in achieving highly reliable and accurate position solutions fulfilling the requirements of Location-Based Services (LBS) pedestrians’ applications. The new strategy is divided into two main parts. The first part integrates the available positioning technology within the surrounding LBS application context by introducing an adaptive LBS framework. The context can be described as a group of factors affecting the application behaviour; this includes environmental states, available resources and user preferences. The proposed adaptive framework consists of several stages, such as defining the contextual factors that have a direct effect on the positioning performance, identifying preliminary positioning performance requirements associated with different LBS application groups, and introducing an intelligent positioning services selection function. The second part of this work involves the design and development of a novel positioning model that is responsible for delivering highly reliable, accurate and precise position solutions to LBS users. This new model is based on the single frequency GPS Standard Positioning Service (SPS). Additionally, it is incorporated within the adaptive LBS framework while providing the position solutions, in which all identified contextual factors and application requirements are accounted. The positioning model operates over a client-server architecture including two main components, described as the Localisation Server (LS) and the Mobile Unit (MU). Hybrid functional approaches were developed at both components consisting of several processing procedures allowing the positioning model to operate in two position determination modes. Stand-alone mode is used if enough navigation information was available at the MU using its local positioning device (GPS/EGNOS receiver). Otherwise, server-based mode is utilised, in which the LS intervenes and starts providing the required position solutions. At the LS, multiple sources of GPS augmentation services were received using the Internet as the sole augmentation data transportation medium. The augmentation data was then processed and integrated for the purpose of guaranteeing the availability of valid and reliable information required for the provision of accurate and precise position solutions. Two main advanced position computation methods were developed at the LS, described as coordinate domain and raw domain. The positioning model was experimentally evaluated. According to the reported results, the LS through the developed position computation methods, was able to provide position samples with an accuracy of less than 2 meters, with high precision at 95% confidence level; this was achieved in urban, rural, and open space (clear satellite view) navigation environments. Additionally, the integrity of the position solutions was guaranteed in such environments during more than 90% of the navigation time, taking into consideration the identified integrity thresholds (Horizontal Alert Limits (HAL)=11 m). This positioning performance has outperformed the existing GPS/EGNOS service which was implemented at the MU in all scenarios and environments. In addition, utilising a simulation evaluation facility the developed positioning model performance was quantified with reference to a hybrid positioning service that will be offered by future Galileo Open Service (OS) along with GPS/EGNOS. Using the statistical t-test, it was concluded that there is no significant difference in terms of the position samples’ accuracy achieved from the developed positioning model and the hybrid system at a particular navigation environment described as rural area. The p-value was 0.08 and the level of significance used was 0.05. However, a significant difference in terms of the service integrity for the advantage of the hybrid system was experienced in all remaining scenarios and environments more especially the urban areas due to surrounding obstacles and conditions.