Μελέτη, αναλυτική προδιαγραφή και υλοποίηση ενός συστήματος διαχείρισης πληροφορίας περιβάλλοντος για διεισδυτικές εφαρμογές
Ubiquitous and pervasive computing environments form a significant vision for the future of computing and aim to replace the way computing is currently involved in everyday life. These environments will offer computing capabilities integrated in almost everything, in such a manner that the will be able to respond to the needs of each person. Networks, services and devices will cooperate via transparent and unobtrusive processes to form a unified virtual mesh that is available everywhere and can support the activities of people within it. These processes should require minimal user intervention and be able to adapt dynamically to the needs of each person in an intelligent manner, so that they will be actually useful instead of troublesome for the end users. To support these operations, such an environment must provide continuous and uninterruptible access to the network and its services. It is thus necessary to automatically perform processes such as discovery, location, configuration and management of the various entities within the environment, so that these entities can work effectively towards catering for the user’s needs. These operations must not be interrupted as the location of users within the environment changes. Such an environment should also simplify the management and dispersion of intelligence and acquired knowledge. Therefore, it should provide context awareness as a service to executing processes. Consequently such a system needs to collect, evaluate, classify and sort information regarding the context of all physical and virtual entities within the environment. This process is called context management and is necessary so that the system can leverage the information collected and the knowledge acquired in order to extract higher level information and be able to provide personalized services to each user. Finally, since a significant part of the collected information may involve personal and sensitive data about users, it is very important to provide mechanisms for ensuring the security of such information and the privacy of the users. This paper presents the requirements analysis, design and implementation of a Context Management system. SED (Sensor Event Detector) provides a robust mechanism for collecting data from various sensors, semantic uplifting and context inference, and managing the context state within a smart environment. SED was designed in such a way so that it can be used in a smart environment in order to allow the provision of personalized services to users according to their preferences, location, as well as temporal and physical conditions and other context information collected.