Thesis Abstract - Take One

Normally, you don't write the Abstract until you have actually done the thesis. Go figure. But I have decided to give it a jab and put something down for the time-being. Obviously, I will be returning every now and again to shape it up, but this is what I have in mind for now:

Abstract -- The next generation network (NGN), as specified by the International Telecommunication Union (ITU), seeks to guarantee that the network of the future will provide users of telecommunication systems with seamless access across disparate networks, to the applications and services they need and use in everyday life. The IP Multimedia Subsystem (IMS) as defined by standards bodies such as the 3GPP (3rd Generation Partnership Project), ETSI (European Telecommunication Standards Institute) and others, is an important part of the NGN, particularly from the perspective of providing the services architecture that executes and delivers services based on user requests. The decomposition of these services into their constituent parts provides IMS engineers with the advantage of code re-usability and a more flexible platform upon which to develop complex services. Work in this area has led to the definition (outside of the standards) of a service capability interaction manager (SCIM) and a service broker which are responsible for handling the interactions between service components and managing access and third party exposure to these components respectively. The aim of the work related in this thesis has been to investigate in detail the service architecture of the IMS and to identify and validate the shortcomings, if any, of the current architecture for the delivery of complex services through service orchestration. The work applies only to the composition of services that are hosted on SIP (Session Initiation Protocol) application servers which is one of the three target environments specified by IMS. In agreement with related work, this thesis finds that the current architecture is not fully supportive of these functions without some extra interventions. A novel architecture that extends the currently proposed model is subsequently presented using the Mobicents platform which is based on the JAIN SLEE standard that, unlike related work, re-uses existing IMS standards and mechanisms and is compared and contrasted with contemporary designs such as Parlay X, GSMA One API and OMA NGSI.