Equalisation of satellite mobile channels with neural network techniques

Article type: Research Article

Authors: Bouchired, Steven^{; ;} | Roviras, Daniel | Castanié, Francis

Affiliations: TéSA, 17 bis rue Riquet, 31000 Toulouse, France | CRIL TECHNOLOGY, Tersud, 5 avenue Marcel Dassault, 31500 Toulouse, France Tel.: +33 5 61 58 84 46; Fax: +33 5 61 58 82 37; E‐mail: roviras@len7.enseeiht.fr

Note: [] Corresponding author.

Abstract: The use of non‐linear amplifiers near saturation in satellite channels causes severe distortion of the transmitted signal. These distortions make the matter of power efficiency prevail over spectral efficiency in satellite communication systems. Nowadays, the use of satellite segments for applications requiring ever increasing data rates puts this principle in question. The present article proposes to fight non‐linear distortion caused by satellite amplifiers by resorting to simple neural network equalisation devices in the receiving earth terminals. Several neural network equalisers are introduced, and applied to a satellite Universal Mobile Telecommunication System (S‐UMTS) channel model. They are shown not only to outperform conventional equalisation techniques like the linear transversal equaliser, but also to have better performance than the non‐linear Volterra equaliser. The performance improvement is all the more sensitive as the modulation scheme used is severely distorted. The case of 16‐QAM transmission over a mobile satellite link is studied in details.

Journal: Space Communications, vol. 15, no. 4, pp. 209-220, 1999