Affiliations: Faculty of Electrical Engineering and Computing, University of Zagreb, Unska 3, HR-10000 Zagreb, Croatia
Correspondence:
[*]
Corresponding author: Faculty of Electrical Engineering and Computing, University of Zagreb, Unska 3, HR-10000 Zagreb, Croatia. Tel.: +385 1 612 98 02; Fax: +385 1 612 98 32; E-mail: ignac.lovrek@fer.hr
Abstract: The paper deals with mobility management in the all-mobile network where mobile and intelligent software agents represent the users in the network and support terminal, user and service mobility. A mobility management model, suitable for the agent’s migration between the user terminals and network nodes according to the tasks defined by its user and user’s location, is discussed. The elaborated knowledge-based mobility management is defined as one that reduces location uncertainty using knowledge about user’s favourite locations, habitual duration of stay and routes. The concept of mobility events and movement knowledge representation for intelligent search for the user are proposed. Case studies describing mobility management approach based on the “most likely locations in the most likely order” principle are included. The results of the model analysis, including discussion about regular and random user’s movement, as well as location and path knowledge, are presented.
