Topics in signal design and detection for broadband satellite communications

This thesis deals with different aspects of modern satellite communications. Satellite-based telecom systems provide the best and quickest solution to reducing the digital divide in less-favored areas owing to their wide geographical coverage and the speed and ease of deployment of terminal equipment. However, satellite system must be efficient and cost effective, and capable of full inter-working with state-of-the-art terrestrial broadband networks. The first part of this thesis presents a novel two-way broadband satellite architecture. The system provides TV-Centric triple play services to actual end-users in the home employing DVB-S2 and DVB-RCS technologies in the forward and the return link, respectively, together with a newly dynamic resource allocation technique based on the adoption of the Adaptive Coding & Modulation (ACM) functionality together with a QoS policy. The second part analyses alternative satellite transmission schemes reporting benefits and proposing solution to tackle critical issues for two attractive solutions: OFDM and CPM. OFDM technologies are largely used in terrestrial networks but they are not very popular in satellite systems due to the large peak-to-average power ratio. The multicarrier nonlinear channel capacity is derived together with a predistortion algorithm, based on the fixed point iteration (FPI), to mitigate the nonlinear distortions. The performance of a turbo-coded FPI-predistorted OFDM system in the nonlinear channel is also presented. Continuous-phase modulations (CPMs) have the attractive properties of constant envelope and bandwidth efficiency, that make this family the most suitable modulation in the satellite field. In spite of these favourable features the CPMs have found no application until now, except for very simple schemes as MSK, because of the implementation complexity of the detector, and of synchronization problems. Two synchronization algorithms, a cyclostationary-based timing estimator and a soft-based phase recovery method for CPMs, are derived and the Root-Mean Square Estimation Error is shown to assess the performance. A spectral analysis identifying a subclass of CPM signals that are suited for future global navigation satellite systems in the C-band region is also presented.