A Neural Network-based Equalizer for Synchronous Mc- Cdma Variable-bit-rate Transmissions over Leo Satellite Channels

This work is aimed at investigating the use of neural network-based adaptive receivers for synchronous variable bit-rate Multi-Carrier CDMA systems transmitting over low-earth-orbit (LEO) satellite channels. It is known by literature that MC-CDMA techniques are much more resilient with respect to multi-user interference effects in multipath fading channels than single-carrier DS/CDMA ones. Moreover, MC-CDMA exhibits a natural capability to deliver multirate services simply by assigning to each user a variable-cardinality set of subcarriers. However some problems should be faced related to the correct channel estimation and equalisation that are necessary to MC-CDMA schemes for providing acceptable BER results. The proposed neural-network (NN) based MC-CDMA receiver allows to overcome the limitations involved by the conventional equalisation methods, reaching a significant performance improvement of the considered transmission system.