Resumen de Multi-beam satellite resource allocation optimization for beam hopping transmission
Lei Jiang
Multi-beam satellite systems have been studied a lot in the last ten years. They have many promising features like power gain, interference reduction, high flexibility to adapt the asymmetric traffic distribution, and the improvement of the system capacity compared with single-beam systems. In multi-beam satellite systems, the beamforming antenna can generate a number of spot beams over the coverage area. However, each beam will compete with others for resources to achieve satisfactory communication. This is due to the fact that the traffic demand is potentially highly asymmetrical throughout the satellite coverage. Therefore, in order to achieve a good match between offered and requested traffic, the satellite requires a certain degree of flexibility in allocating power, bandwidth and time-slot resources. Current multibeam satellite systems with regular frequency reuse and uniform power allocation can not satisfy these increasing requirements, which motivate us to investigate new transmission schemes to replace the current ones. In this dissertation, we first propose a novel system design, flexible system, which is an extension of current multi-beam systems. It is characterized by the non-regular frequency reuse and the flexibility in bandwidth and power allocation. Then, the Beam Hopping (BH) system is proposed to evaluate the performance improvement with the flexibility in time/space and power domain. As we know, the flexible system and BH system operate in frequency and time/space domain, respectively. In order to know which domain shows the best overall performance, we propose a novel formulation of the Signal-to-Interference plus Noise Ratio (SINR) which allows us to prove the time/frequency duality of these two schemes. Furthermore, to efficiently utilize the satellite resources (e.g., power and bandwidth), we propose two capacity optimization approaches subject to per-beam SINR constraints. Moreover, due to the realistic implementation, a general methodology is formulated including the technological constraints, which prevent the two systems dual of each other (named as technological gap). The Shannon capacity (upper bound) and the state-of-art Modulation and Coding (MODCOD) are analyzed in order to quantify the gap and evaluate the performance of the two candidate schemes. Comparing with the current conventional systems, simulation results show significant improvements in terms of power gain, spectral efficiency and traffic matching ratio. They also show that the BH system is less complex design and outperforms the flexible system specially for non-real time services. This part of the Ph.D. work supported by an ESA-funded project on next generation system of "Beam Hopping Techniques for Multi-beam Satellite Systems". This research is in close collaboration with the leading space industry (e.g. INDRA, MDA) and space research institutions (e.g., ESA, DLR (German Space Agency)). In addition, we extend the work to mobile environments (e.g., railway scenario). Since the current air interface standards (e.g., DVB-S2/RCS) lack of specification for mobile scenarios, a new Fade Mitigation Technique (FMT), i.e., Link Layer Forward Error Correction (LL-FEC) is introduced as a fading countermeasure for DVB-S2/RCS in mobile environments. This part of the work points out that LL-FEC can overcome the deep fading in mobile satellite scenarios (e.g. railway) by optimizing the FEC codes (e.g. Reed-Solomon and Raptor codes). We have to note that such air interface standards might need change to adapt to the new proposed systems: flexible and BH. However, the methodology presented is also applicable. We further investigate the secure communication of multibeam satellite systems by using the system model developed in the BH project. The physical (PHY) layer security technique is investigated to protect the broadcasted data and make it impossible to be wiretapped. A novel multi-beam satellite system is designed to minimize the transmit power under the constraints of the individual secrecy rate requested per user. The main contributions of this Ph.D. dissertation can be summarized as: a. We study the resource allocation optimization in multi-domain (frequency, time, space and power) for multi-beam satellite systems. We prove that the system performance can be increased by dynamically adapting the resource allocation to the characteristics of the system, e.g., traffic requested by the terminal; b. Theoretical studies and simulations prove that the proposed novel transmission schemes perform better than the current system design in terms of power gain, spectral efficiency, etc.. In addition, BH system turns out to show a less complex design and superior performance than the flexible system; c. We prove the theoretical duality between the flexible and BH systems, which works in frequency domain and time domain, respectively. Moreover, we develop a general methodology to conclude the technological constraints due to realistic implementation, obtain the main factors that prevent the two technologies dual of each other in practice, and formulate the technological gap between flexible and BH systems; d. We extend the work to mobile scenarios and prove that LL-FEC is applicable for mobile satellite systems (e.g., railway) to compensate the fade due to the mobility by optimizing the FEC codes (Reed-Solomon and Raptor codes). The results show that Multiple Protocol Encapsulation Inter-burst FEC (MPE-IFEC) and extended MPE-FEC with Raptor codes - as finally specified in DVB Return Channel via Satellite for Mobile Scenario (DVB-RCS+M) - consistently perform better than other LL-FEC schemes for mobile scenarios; e. We point out that how to change the signalling of current version of standards (e.g., DVB-S2/RCS+M) in order to allow achievable performance in the mobile scenarios. The proposal has been finally adopted by the DVB-RCS+M standard. f. We finally make use of our developed system models to investigate whether the multibeam scenario allows the use of PHY layer security, a very valuable feature that would broaden multibeam satellite applications. We prove that our models are directly applicable for the study of PHY layer security in terms of joint optimization of power control and beamforming for the BH payload. Moreover, the proposed algorithm can ensure the minimum power consumption subject to the individual secrecy rate requested per user. Based on the work of the Ph.D., three journal papers and eleven international conference papers have been published, and these publications systematically cover all the contributions of this doctoral thesis work.
