The ultimate aim of the information revolution is to ensure that everyone can have access to information, anytime, anywhere - at reasonable cost. To achieve this goal, the emerging satellite technology gives a new perspective for a universal access to the broadband infrastructure, potentially alleviating the prohibitive cost of serving every user by terrestrial digital networks. More than any other digital transmission technology, satellite communication is predestined, in combination with optical transmission, mobile communication and high-speed microelectronics, to introduce "integrated broadband communication" even to the non-professional user. Furthermore, by ideally complementing the terrestrial networking infrastructure and because of the possibility to integrate fast global telecommunication services into one single network, satellite communication appears as the key technology to make the vision of "global communication" serving the emerging "information society" a reality.
Satellite communication is now at a crucial juncture, in that new, advanced, satellite system designs are dramatically changing the cost-structure of the technology. Satellite-enabled personal communication services including two-way voice, fax, data, video, teleconferencing, multicasting and broadcasting are becoming increasingly affordable to residential, but also and more importantly, mobile users. Societal globalisation processes are expected to significantly benefit from low-cost, satellite-enabled international roaming.
Besides telecommunication technology developments, many research and development activities in distributed and multimedia computing have evolved over the last few years. As opposed to conventional computing and data communications operating with alphanumeric data, multimedia computing due to the type and nature of multimedia data exhibits whole new classes of system requirements with respect with capturing, storing, streaming, transmission, synchronisation and presentation, commonly referred to as Quality of Service (QoS) requirements. First approaches in coping with those requirements on the network traffic level have led to the evolution of an ATM-based Broadband Integrated Services Digital Network (B-ISDN), which can not only support high transmission rates, but can also allow different applications or multimedia streams to be transmitted simultaneously in an integrated manner.
Terrestrial and satellite links will provide ubiquitous access to a plethora of multimedia services over B-ISDN and to the Internet. Geostationary satellites offer several unique benefits to "global telecommunication": ubiquitous wide-area coverage, network flexibility, including broadcast and multipoint-to-multipoint capabilities, as well as rapid network set-up and deployment. On the other hand, for areas with a fibre optic infrastructure, the terrestrial networks are often sufficient to provide access to fixed residential users and companies.
Communications between mobile users is increasing and it will constitute a large market at the beginning of the next millennium. With increased deployment of wireless technologies, it appears that a very promising solution is the installation of Wireless Local Loops (WLL) to provide network access at low cost. These access methods can also be combined with other existing terrestrial networks.
In order to efficiently handle multimedia applications across networks in general, and through the air interface in particular, appropriate network capacity management mechanisms are necessary. Network capacity planning is the practice of anticipating the future needs of networked users, and designing and managing the network resources accordingly. Two key issues play a dominating role, first the characterisation of the type and quality of service a user expects from the network, and second, the characterisation of the type and quantity of load a user generates into the network.
In this work, the accurate characterisation of these parameters will be attempted by instrumentation, modelling and analysis of user behaviour, and by developing models of network load and traffic as induced by user activity in future multimedia applications with the goal of supporting the design and the capacity management of future networks. As the data exchange for multimedia services requires very high capacity networks, it is useful to consider the operational characteristics of a High-Performance Computing Network (HPCN), particularly in the context of the symbiosis of terrestrial optical transmission and communication satellites. The proliferation of a network capacity planning workbench with the ability to investigate future networking scenarios, i.e. the consideration of a manifold of heterogeneous groups of individuals with different user behaviour, operating a variety of multimedia applications in a global network, will help the European telecommunication industry to early identify complexity, scale and technology trends of broadband networks.
Due to the possibility of reducing the time-to-react to approaching telecommunication trends, this work will make European vendors more competitive in the global broadband satellite communication market. It is of utmost importance also for the integration of the individual citizen into the upcoming "information society" to provide communication facilities to "everyone", and to deliver seamless Quality of Service according to individual user needs. Capacity planning for future networks serving future users must be started now.
Future network capacity planning activities like sizing the network, allocating dedicated bandwidth, bounding latency, or to guarantee end-to-end QoS, but also detailed network performance analysis like the evaluation of protocol designs, traffic shaping or the investigation of routing policies, involves system and user modelling, and the respective model evaluation.
The proposed work will develop network system and network traffic models based upon profiles of user behaviour. The study will establish insight on how new multimedia and network computing technologies (such as e.g. Java, CORBA, VRML) will influence network performance, and maybe alter user behaviour. The mapping and implications of QoS parameters at various system levels (network-perceived QoS, user-perceived QoS) will be investigated.
The means for network model evaluation are large-scale discrete event simulations. Appropriate approaches to the network simulation scalability problem is the use of parallel and distributed simulation techniques, where the simulation task is decomposed and assigned to a set of processors which execute subtasks concurrently in a coordinated way. A complementary solution is to slim down the simulation by abstracting out details under controlled risk, thus enabling modular, hierarchical simulations at arbitrary levels of accuracy. The simulation results will be compared with measurement results from a test-bed to verify the simulation models.
Investigations will encompass the characteristics of Low-Earth-Orbit (LEO) and Geostationary-Orbit (GEO) satellite constellations, yielding modular satellite network models. The user behaviour and application characteristic models will then be integrated in order to carry out simulations to verify that the application receives the level of service it requires while efficiently using the capacity of the air interface. The effects of the multiple access scheme (TDMA, FDMA, CDMA) and of the satellite link characteristics on the applications will be investigated to optimise the utilisation of network capacity.
The project will involve six non-affiliated participants from four European Union (EU) Member countries. It is a highly targeted project involving only partners with complementary know-how that will make the project successful, incorporating: