Modelling and Measurement of the Land Mobile Satellite MIMO Radio Propagation Channel.

Three crucial factors in land mobile satellite (LMS) communication systems are the quality of service (QoS), spectral efficiency and cost. The QoS in a LMS system often suffers due to high link path loss due to the vast distances covered, signal shadowing and blockage, and a high link delay. Spectral efficiency can also be fairly low in LMS systems due to small received signal to noise ratios disabling the adoption of high order modulation techniques. Setup cost is also a major factor influencing the business case for LMS communication systems, which often makes single satellite systems more attractive than multiple satellite constellations. This thesis advances a technique for increasing QoS and spectral efficiency, without any increase in total transmit power, antenna gain or bandwidth by using multiple-input multiple-output (MIMO) techniques, which is then investigated thoroughly by theoretical and experimental means. To investigate the benefit from satellite-MIMO techniques, work began with the design of a satellite-MIMO physical-statistical channel model, which enables computer simulation of the LMS-MIMO channel. Although the single-input single-output (SISO) channel, a subset of the MIMO channel was validated partially to previously published measurement data, there was no measurement campaign data or literature that informed about the LMS-MIMO channel. The model was used to carry out an initial estimation of capacity and diversity gain available from LMS-MIMO systems. Work moved onto the wideband characterization of the satellite-MIMO channel. A measurement campaign was carried out in Guildford, U.K., where a terrestrially based artificial platform, representing two low elevation satellites in a cluster communicated with a mobile van, situated in three environments: tree-lined road, suburban and urban. Each emulated satellite contained a right and left hand circularly polarized (R/LHCP) directional antenna mounted next to each other. The vehicle contained four spatially separated omnidirectional antennas: two RHCP and two LHCP. A complete analysis was performed on the dual polarization single satellite 2x2 MIMO channel data. The data was characterized in terms of narrowband and wideband first and second order statistics for both large and small scale channel fading. The large and small scale correlation statistics were also extracted from the data across the delay and MIMO domains. The data was used to estimate the capacity and diversity gain that could be achieved from a low elevation LMS-MIMO channel. This analysis also resulted in the derivation of empirical-statistical narrowband and wideband satellite-MIMO channel models, which have been validated against the experimental measurement data. The models can be used by the global research community to help design advanced PHY and MAC layer techniques, and some interest has already been shown.