Investigations of Channel Capacity Loss in LEO Satellite Systems Using Phased Array Beamforming Antennas

In LEO communication systems phased array beamforming antennas are used extensively on terminal and satellite side. The main purpose of such antennas is electrical beam steering without using any mechanical devices. With this, the beam has to follow the overflight trajectory. However, phased array antennas only have a limited set of discrete beam steering directions, which is the main drawback. The antenna controller has to select the best-suited beam from the given pointing vector set at any time of the flyover scenario. This always results in non-optimal antenna steering. On the receiver side, this has the effect of a strongly fluctuating reception power. This paper aims to show how discrete beam pointing can reduce the channel capacity of LEO satellite systems by up to 10 this effect in the early planning phase of LEO satellite systems, because in this project stage system engineers have to determine the feasibility and establish initial coarse system environment parameters. For this purpose the antenna gain variation is introduced, as a kind of statistical channel fading. Additionally, we propose the loss in spectral efficiency as a suitable performance indicator. This allows us to make early reliable assessments for orbit altitude, inclination angle, transmission power, etc. Furthermore, we outline those design parameters which have a significant impact on the LEO channel capacity. Moreover, we show that the loss of ergodic spectral efficiency also occurs in realistic scenarios because typical modulation and code rate formats are operating very close to the Shannon limit. This drawback can be surmounted by several mitigation methods, which are presented here. These methods are based on modulation and code rate schemes in conjunction with channel prediction algorithms. Each algorithm is analyzed and compared in detail. Monte-Carlo simulations demonstrate the benefit of a neural network or the presented memory-efficient Fourier series predictor in channel prediction. In comparison to typical terrestrial channel predictors, the capacity loss is much lower.