MIMO Satellite Communications: A New Application of Multi-Antenna Technologies and Network Coding under Line-of-Sight Conditions
Final Report Abstract
This final summarizes our main results in the field of Multiple Input Multiple Output (MIMO) systems for Satellite Communications in the period from March 2010 to February 2012. The proposed work schedule consisted of 6 Work Packages (WP). Our actual activity concentrated on the WPs “Channel Models”, “Capacity Considerations”, “Sensitivity Analysis” and “MIMO Precoding”. Whereas MIMO systems are nowadays deeply widespread in mobile communications, their implementation in the satellite communications is not yet established. The main reason is that the characteristics of the satellite communication channel are deeply different from the ones of radiomobile terrestrial channels, and this does not allow a direct transposition of the techniques used for the latter. Due to hardware and power limitations in the satellites, transmission strategies must limit the required processing and take into account the non-linearities induced by high power amplifiers driven near saturation for power efficiency. Moreover, the satellite channel usually exhibits a dominant line-of-sight (LOS) signal component, whereas usual MIMO systems take advantage of scattered (NLOS) components. This latter constraint may be overcome from a theoretical point of view by considering that the capacity gain achievable through MIMO techniques (known as “spatial multiplexing”) actually depends only on the rank of the MIMO channel matrix, and that any degrees of freedom which allow to achieve a high-rank matrix may be exploited. Whereas terrestrial systems take advantage of the statistical uncorrelation of scattered signal components, satellite systems (and generally speaking all wireless systems with a strong LOS component) may instead exploit the geometrical displacement of antenna elements to produce a maximum-rank channel matrix. Early work at our institute dealt with MIMO LOS channels for indoor applications; later works applied the same principles to the MIMO satelite channel and derived a general criterion for the antenna positions which maximizes the channel capacity and proposed some application scenarios, under the hypothesis that the propagation model is fully described by its deterministic LOS component. The correct displacement of antennas according to this criterion becomes the key to achieve spatial multiplexing also in the satellite scenario. This was the starting point for the work covered by this report. We outline here our main new results. • A MIMO satellite channel geometrically optimized in the ideal conditions of free space LOS propagation still offers maximum multiplexing gain even in presence of severe weather effects, since any phase disturbances which might affect the signals traveling through the troposphere do not degrade the orthogonal structure of the channel matrix. Furthermore, due to site diversity, MIMO satellite systems would offer as a side effect outstanding improvements in the system reliability. • In a dual-satellite scenario, the capacity loss due to imperfect positioning of antennas of the ground station can be mitigated by employing Uniform Circular Arrays (UCA), instead of linear arrays. A basic 3-elements UCA would make the achievable capacity practically independent of the terminal orientation angle and its size would be suited to a vehicle rooftop installation. We derive this result through an analytical optimization of the geometric parameters followed by numerical simulations, as a closed-form solution is not affordable. • If the transmitter knows the MIMO channel, precoding techniques can be applied to improve the reliability of the link when the channel is not perfectly orthogonal (e.g. antennas mispositioning, movements of the satellites in their station keeping box, . . . ). Because a satellite link must achieve a given Quality Of Service (QoS) (e.g. Bit Error Rate constraint), we considered the joint design of bit loading and input-output mutual information maximizing precoding strategies. A per-antenna power constraint was used for the precoder design since each transmit antenna is usually fed with its own power amplifier. Link level simulations of these approaches showed that, under a BER constraint, a high throughput can be ensured even for non orthogonal MIMO channels paving the way for robust MIMO satellite transmissions. • The basic MIMO precoding scheme based on the singular value decomposition (SVD) of the MIMO channel matrix would be quite robust towards amplifier nonlinearities, in spite of the high peak-to-average power ratio which is introduced by linar precoding. Nevertheless, other robust precoding schemes should be investigated, which would be more suitable for practical implementation. • Merging spatial multiplexing and network coding into a MIMO two-way relaying scheme would allow 25% bandwidth saving compared to the state of the art. This is an early concept, which would need a separate research project for further development. With regard to the WP “Channel Models”, we had to renounce to actual field measurements, as the hardware currently available on satellites does not obviously support MIMO transmission. Finally, we would like to recall here that the focus of this project is on the practical feasibility of MIMO techniques in the concrete context of satellite communications. For this reasons, a theoretical capacity analysis of general relay channels, which was suggested by the reviewers, has not been carried out. This topic would be better suitable for a more theoretical research project.
Publications
- ”On the Capacity Degradation in Broadband MIMO Satellite Downlinks with Atmospheric Impairments”, IEEE ICC 2010 - General Symposium on Selected Areas in Communications, Cape Town, South Africa, May 2010
A. Knopp, R.T. Schwarz, B. Lankl
(See online at https://doi.org/10.1109/ICC.2010.5502507) - ”Uniform Circular Arrays: the Key to Optimum Channel Capacity in Mobile MIMO Satellite Links”, 5th Advanced Satellite Mobile Systems Conference (ASMS) and the 11th Signal Processing for Space Communications Workshop (SPSC), Cagliari, Italy, 13-15 Sept. 2010
V. Dantona, R.T. Schwarz, A. Knopp and B. Lankl
(See online at https://doi.org/10.1109/ASMS-SPSC.2010.5586873) - “Impact of nonlinear power amplifiers on the performance of precoded MIMO satellite systems”, ESTEL 2012, IEEE First AESS European Conference on Satellite Telecommunications, 2-5 Oct. 2012. - Awarded as “Best Student Paper”
V. Dantona, T. Delamotte, G. Bauch, B. Lankl
(See online at https://doi.org/10.1109/ESTEL.2012.6400133) - “Robust Power Allocation and Bit Loading for Fixed Wireless MIMO Channels”, SCC 2013, ITG Conference on Systems, Communications and Coding, Munich, Germany, 21-24 January 2013
T. Delamotte, V. Dantona, G. Bauch, B. Lankl
