Beam Hopping in Multi-Beam Broadband Satellite Systems: System Performance and Payload Architecture Analysis

Abstract

When considering the inherent uncertainty of the broadband communication market evolution as well as the spatial non-uniform distribution of capacity requests through the satellite coverage, flexibility to adapt to different traffic configurations is a key requirement for future satellite systems to maintain their competitiveness. Beam hopping solutions appear as good candidates to match system offered capacity with traffic demand. Although the introduction of flexible bandwidth and power allocation to beams brings some benefits in terms of demand satisfaction for hot spots with respect to traditional systems with regular frequency reuse and equal power to beam allocation, utilization of beam hopping techniques offers an additional mean to cope with irregular and time variant traffic distributions. In this paper the forward link of a multibeam transparent satellite system for providing broadband Internet access to users sparse over the coverage area is analyzed. The utilization of Adaptive Coding and Modulation (ACM) techniques is considered, according to the recently approved DVB-S2 standard. The advantage in terms of system performance brought by the utilization of beam hopping techniques is assessed. In this type of system, only a few beams of the overall coverage are selected at each time-slot for transmission. This set of illuminated beams changes on a slot-by-slot basis over the duration of a repetition window. This creates a time and spatial transmission plan that repeats periodically with a given duty cycle. In the present contribution the problem of optimizing the beam hopping transmission plan in order to best match the traffic demand is tackled. Genetic algorithms appear as the most suited optimization tool because of their global optimization nature, their capability of working with a large number of optimization parameters and their suitability to accept a discontinuous figure of merit function. Extensive simulations have been carried out to adapt genetic algorithms functions and parameters to the problem under investigation. The target has been optimizing the beam allocation in the transmission plan to maximize system capacity. System performance is assessed considering the traffic demand distribution over Europe foreseen for broadband satellite services in 2015, as derived by recent market analysis studies. A throughput increase of about 30% compared to a system with regular power and bandwidth allocation capabilities is achieved. The impact of a beam hopping type of system at payload level is then investigated. In particular, efficient ways to implement beam hopping combining payload flexibility with affordable complexity are investigated. Both Direct Radiating Array (DRA) and Array Fed Reflector (AFR) solutions are considered.