A Unified Scalable Model of User Localisation with Uncertainty Awareness for Large-scale Pervasive Environments

Abstract

Localisation has become a standard feature in many mobile applications. Numerous techniques for both indoor and outdoor location tracking are available today, providing a diversity of ways positioning information can be delivered to a mobile application (e.g., a location-based service). Such factors as the variation of precision over time and covered areas or the difference in quality and reliability make the adoption of several techniques for one application cumbersome. This work presents an approach that models the capabilities of localisation systems and then uses this model to build a unified view on localisation, with special attention paid to uncertainty coming from different localisation conditions and its presentation to the user. We discuss technical considerations, challenges and issues of the approach and report about a user study on users' acceptance of the suggested behaviour of an application based on the approach. The results of the study showed the feasibility of the approach and revealed users' preference towards automatic but yet informed changes they experienced while using the application.

Location Information Quality: A Review

“…Baus and Kray [ 6 ] define four different higher level qualitative categories for positional information: precise information, unprecise information, no information and false information. Aksenov et al [ 7 ] present a middleware that encompasses different localization systems. To describe the parameters of each system, the authors use update rate and precision.…”

“…While the above aspects capture spatial aspects of positional information, we also identified two main aspects of the temporal dimension: update rate and recency. The update rate [ 7 ], (also known as frequency [ 13 ] or sample interval [ 17 ]) describes how often a sensor measures a property. We model it as how often a new location measurement is acquired or as how many times per second the location is sampled.…”

O brother, where art thou located?: raising awareness of variability in location tracking for users of location-based pervasive applications

“…Also, within the context of the study described in section 4, we assigned the walking speed V=1.2 m/s, which is around the lower end of the scale of the average walking speed A request of the first type is triggered when an update (L' in Figure 3) on a user's location has been received and needs to be communicated. This update is first processed accordingly by the Location Processing Component (introduced and described in (Aksenov et al 2011); hereafter abbreviated to LPC), which updates the ontology and then sends the updated instance to the mapping table (L to UI Mapping Table in Figure 3). The mapping table is a collection of rules that determine which subset of visualisation patterns corresponds to the attributes of the received location update.…”

“…Our previous work (Aksenov et al 2011) also falls into the category of research initiatives on the representation and management of uncertainty about location. We introduced an ontology that models the interiors of localisation systems and used that model to build a unified view on localisation on a large-scale, with a separate focus on uncertainty and variability in location sensing and their presentation to the user.…”