The growing interest in multimedia fixed applications calls for the development of point-to-point satellite systems capable of providing high-speed links at a competitive price. In order to meet this goal, satellite systems need to significantly increase their overall throughput with respect to current state-of-art.
Future multimedia broadband satellite systems will thus rely on:
- Large number of beams,
- ACM techniques to counteract fading / interference.
This study was motivated by the need to investigate new techniques for capacity improvement. In particular, the use of satellite systems with large number of beams and operation at higher frequency makes the system performance interference limited. Hence, techniques which go beyond ACM in mitigating interference and improving system throughput were the objectives of this work.
The main objectives of the Phase I of the present study were the exploration of new Multi-User-Detection (MUD) techniques able to improve the spectral efficiency, and hence the system capacity, of future broadband satellite systems for fixed communications.
In particular, the proposed techniques shall provide further improvements with respect to state-art ACM-enabled systems. The reference system for the new techniques evaluation was a transparent multi-star satellite network operating at high frequency band (Ka or higher) and availing of an antenna system able to generate a large number of beams (about 100 or even more).
Also within the scope of this study was the preliminary investigation of the applicability to the mobile-satellite communication field of the techniques already proposed and investigated in the frame of applicability to fixed satellite systems. A further topic of investigation was the analysis of interference mitigation in the frame of a spread Aloha system.
In Phase II the main objective was the development of a SW tool able to make detailed performance assessment of the most attractive of the proposed solutions. Performance assessment was required to be done not only at physical layer level but also at system level.
The most important issues addressed in the study were:
- The improvement of the FL of a multi-beam satellite system through the exploitation of precoding techniques at the GW side.
- The improvement of the RL of a multi-beam satellite system using a DVB-RCS like TDMA access. Preferred techniques for this scenario were the spatial MMSE processing, eventually followed by SIC (Successive Interference cancellation) or Iterative Interference Cancellation (IIC).
The study has investigated a large spectrum of possible techniques to improve the spectral efficiency of current satellite systems. In particular, for the multi-beam fixed scenario spatial processing was shown to be the key for improving the system capacity in a significant matter. It was shown that spatial processing was to be preferred to other techniques, also proposed in the technical literature, like:
- Carrier Pairing. This technique reuses the same bandwidth for both the FL and RL. It was however shown that it performs poorly in a multi-beam system with aggressive frequency reuse. It may however be attractive in a single beam (or very few beams) scenario.
- ACI mitigation through Iterative Interference Cancellation. This technique, which is suitable for the RL only, can only provide marginal gain when compared to spatial processing. Its main attractiveness is its ability to operate also in single beam systems.
Capacity gains which go from 30% on the FL to more than 50 % on the RL were shown to be possible with the considered spatial processing techniques. Even higher gains were shown possible with ad-hoc designed systems.
The following methodology was applied for evaluation of the MUD techniques in the context of fixed broadband satellite systems.
First, a conventional satellite system was designed using state-of-the-art transmission techniques (DVB-S2 in the FL and an ACM-enabled, enhanced DVB-RCS, in the RL).
The satellite system was implementing a European coverage by means of 88 beams of approximately 0.5° beamwidth. The satellite payload was sized taking into account DC power limitations of current satellite platforms.
The system capacity and link availability achievable with the use of the conventional ACM techniques was first evaluated. Then several MUD techniques were considered and their effects on the system capacity and availability of the designed satellite system were assessed. This allowed sorting the proposed techniques based on their effectiveness in the considered scenario.
Within the techniques providing higher system capacity improvements, preference was then given to those with lower complexity. As a matter of fact, selected techniques were all requiring centralized processing at the Gateways this property being very attractive as it avoids any increase in the user terminal complexity.
Techniques based on spatial processing were shown to be the most promising for the considered system scenario.
In particular, for the FL, Precoding was the favoured approach. Several Linear Precoding schemes were evaluated given the relative immaturity of the more advanced Vector Precoding schemes. The main limitation of any Precoding schemes (particularly those linear) is the produced signal envelope fluctuation which makes them only attractive for payload architectures based on active antennas or anyway envisaging multi-carrier HPAs (e.g. MPAs).
For the RL, MMSE beamforming plus Successive Interference Cancellation (MMSE-SIC) or Iterative Interference Cancellation (MMSE-IIC) were shown to be the most attractive techniques. For all spatial processing techniques the issue of channel estimation is of paramount importance. The feasibility and the achievable quality of practical, low-complexity, channel estimation techniques was specifically addressed in the study and workable solutions demonstrated.
The work was divided into two Phases. Phase I had the main objective to list and then sort the most appropriate MUD techniques for next generation transparent broadband satellite systems based on multi-star topology. In particular different classes of solutions were to be found for the Forward-Link and the Reverse-Link of such systems.
Adaptations of some of these techniques to the mobile environment were also considered together with possible enhancement of the classical Spread Aloha technique. In Phase II the effort was directed toward a SW tool development aimed at an in-depth validation of some of the proposed techniques. A simulation campaign using the developed tool was finally performed.
The project has been completed. A Final Report documenting some of the investigated techniques has been produced and the developed SW tool delivered to ESA.