TRANSMISSION OPTIMIZATION FOR HIGH THROUGHPUT SATELLITE SYSTEMS

Demands on broadband data service are increasing dramatically each year. Following terrestrial trends, satellite communication systems have moved from the traditional TV broadcasting to provide interactive broadband services even to urban users. While cellular and land-line networks are mainly designed to deliver broadband services to metropolitan and large urban centers, satellite based solutions have the advantage of covering these demands over a wide geography including rural and remote users. However, to stay competitive with economical terrestrial solutions, it is necessary to reduce the cost per transmitted bit by increasing the capacity of the satellite systems.
The objective of this thesis is to design and develop techniques capable of enhancing the capacity of next generation high throughput satellite systems. Specifically, the thesis focuses on three main topics: 1) Q/V band feeder link design, 2) robust precoding design for multibeam satellite systems, and 3) developing techniques for tackling related optimization problems.
Design of high bandwidth and reliable feeder links is central towards provisioning new services on the user link of a multibeam SatCom system. Towards this, utilization of the Q/V band and an exploitation of multiple gateway as a transmit diversity measure for overcoming severe propagation effects are being considered. In this context, the thesis deals with the design of a feeder link comprising N + P gateways (N active and P redundant gateways). Towards satisfying the desired availability, a novel switching scheme is analyzed and practical aspects such as prediction based switching and switching rate are discussed. Building on this result, an analysis for the N + P scenario leading to a quantification of the end-to-end performance is provided.
On the other hand, frequency reuse in multibeam satellite systems along with precoding techniques can increase the capacity at the user link. Similar to terrestrial communication channels, satellite based communication channels are time-varying and for typical precoding applications, the transmitter needs to know the channel state information (CSI) of the downlink channel.
Due to fluctuations of the phase components, the channel is time-varying resulting in outdated CSI at the transmitter because of the long round trip delay. This thesis studies a robust precoder design framework considering requirements on availability and average signal to interference and noise ratio (SINR). Probabilistic and expectation based approaches are used to formulate the design criteria which are solved using convex optimization tools. The performance of the resulting precoder is evaluated through extensive simulations.
Although a satellite channel is considered, the presented analysis is valid for any vector channel with phase uncertainty.
In general, the precoder design problem can be cast as power minimization problem or max-min fairness problem depending on the objectives and requirements of design. The power minimization problem can typically be formulated as a non-convex quadratically constrained quadratic programming (QCQP) problem and the max-min fairness problem as a fractional quadratic program. These problems are known to be NP-hard in general. In this thesis, the original design problem is transformed to an unconstrained optimizationproblem using the specialized penalty terms. The efficient iterative optimization frameworks are proposed based on a separate optimization of the penalized objective function over its partition of variables at each iteration. Various aspects of the proposed approach including performance of the algorithm and its implementation complexity are studied.
This thesis is made under joint supervision agreement between KTH Royal Institute of Technology, School of Electrical Engineering, Stockholm, Sweden and University of Luxembourg, Luxembourg.