Resource Optimization In Multi-beam Satellites

Abstract

Multi-beam satellite systems offer higher throughput due to frequency reuse andmultiple levels of flexibility and allow the use of smaller earth terminals due to theirhigher directivity. At a high-level multi-beam satellite operation resembles to that ofcellular mobile communications.As the lifetime of a satellite is about 15 years efficient communication satellites musttake into account current as well as future demands. To improve satellite efficiencyand reduce operational risk, design flexibility is imperative. Among others, multibeamsatellites, offer significant flexibility both with regard to coverage area andresources allocation. Coverage area flexibility is achieved by adjusting the position ofthe spots adaptively to communications traffic. Resource allocation flexibility is madepossible by adaptively adjusting the power and bandwidth to the various beamsaccording to traffic.In the present thesis, we focus on resource allocation flexibility offered by multi-beamsatellites. In particular, for a specific satellite payload (number of beams, availablepower, bandwidth etc.,) the first part of the thesis deals with the problem of allocatingpower to different beams to satisfy the demand as closely as possible. Based on amodel relating the power allocated to the data rate offered, we explore the suitabilityof using/ modifying existing optimization algorithms. Among exact methods as wellas approximate methods that are initially considered, a scheme is selected based oncomplexity, convergence, scalability and other issues. The selected algorithm is thenused to optimize the resource allocation. The advantages obtained over other existingmethods are presented.Next, an attempt is made to minimize the DC power consumption of a multi-beamsatellite. In this course, the optimized resource allocation obtained in the first part ofthe thesis is further enhanced, with the aim to minimize the system powerconsumption. This task is deal with employing a multi-objective optimizationalgorithm that aims at meeting the traffic demand as closely as possible whileminimizing the system power consumption. The benefits from using such multipleobjective optimizations in payload design are demonstrated via appropriate simulationresults.