Context Server for Mobile Network-based Services

17.12.2001

Abstract

As the popularity of mobile devices are becoming increasingly popular, new challenges within computing science arises. One of these challenges is how to efficiently use context information to improve applications.

It is essential for efficient use of context, to have a clear picture of what context is. Context can be defined as “any information that can be used to characterize the situation on an entity”. Context aware computing is therefore about how to utilize this information about the entity.

Several attempts have been made to create a general structure for a context aware system. The context toolkit is one such attempt. It provides applications with an infrastructure for developing aware applications.

There are a lot of considerations to take into account when developing a context aware application. A few of these are how to deal with context abstractions, and ambiguity. Other factors include just-in-time information access, how to adapt applications to different contexts, and using context history properly.

A general architecture for context aware applications will provide a lot of benefits, by allowing other applications to be more rapidly integrated and to fully utilize context.

Any context aware application should provide functionality, which is improved by the use of context. It is important when developing applications to take all user groups into consideration. This means providing simple functions for novice users, added functionality for expert users, and designing the system to allow for disabled users as well.

The system developed in this report uses pace, temperature, location and bandwidth as context elements. All elements are used to improve functionality and to provide users with functions which would no have been available without them.

Two major problems was discovered. First that of representing context data. There is no general datastructure for doing this, and so most context aware applications today use ad-hoc structures. The other problem was that of integrating sensors running locally on a mobile device, into a system where sharing of data and sending notifications to users was possible.