Convergence of cellular and broadcast networks from a multi-radio perspective

he demand for cost-efficient provision of mobile multimedia services is faced with the reality of scarce radio resources. The requirement of spectrum efficiency has driven the development of various digital radio technologies — digital audio broadcasting (DAB), digital video broadcasting (DVB), Global System for Mobile Communications (GSM), General Packet Radio System (GPRS), Universal Mobile Telecommunications Systems (UMTS), and so on — optimized for specific services, namely broadcast or mobile communication. However, existing and emerging multimedia services exhibit challenging requirements in terms of asymmetry, interactivity, real time, and multicast communication. This article describes an IP-based multi-radio infrastructure that enables the cooperation of existing radio networks to combine their capabilites to ensure spectrum-efficient provision of high-quality mobile multimedia services. We define multi-radio as the capability of a system to select and combine radio services out of a set of available services to fulfill user service requirements. The selected set might consist in the simplest case of one radio service, offering as closely as possible the required capabilities. In more advanced configurations, a combination of two or even more radio services is required at the same time to do the job. The decision criteria in the selection process are first of all the service requirements of a user, in terms of, say, throughput, delay and cost. However, also network load or other system parameters might be included. But for all these radio service combinations, spectrum is required to enable them. Today, radio networks operate with fixed long-term spectrum allocations. Systems like GSM [1] are available all over the world in exclusively allocated spectrum. The same will hold true for UMTS [2] when it is introduced. However, the required spectrum varies over both time and location. For example, in many regions there is more demand for mobile voice and data communication during business hours than during leisure time. The same relation is often seen between crowded business centers and sparsely populated areas. The actual required spectrum for a radio service changes with time and location. Therefore, we are investigating dynamic allocation of spectrum to radio services instead of the fixed allocation of today [3]. Note that we refer to aggregated user requests when we talk about more demand for a service; individual requests can be demanding, of course, irrespective or time and location. Overall, the multi-radio approach to use spectrum resources more efficiently addresses the cooperation of different radio systems , preferably in a …