The objective of the activity is to design and test a demodulator supported by an artificial neural network for physical layer processing tasks, demodulation, de-mapping and channel decoding. Satcom applications using DVB-S2X standard as well as other air interfaces (e.g. IoT) shall be considered and the most promising one selected for the implementation of a software demodulator. Targeted Improvements: - reduction of the computational complexity up to 85% for some physical layer processing task- reduction of power consumption in the order of 10% at demodulator level- reduction of implementation losses leading to an increase of data rate by up to 25%Description:High data rate communications make use of computationally complex channel decoding schemes to allow error-free links at achievable signal-to-noise ratios. Artificial Neural Networks (ANNs) are proven to provide in general a good approximation to non-linear functions and therefore are considered a technology with potential to bring significant improvement to address physical layer tasks. Besides, recent research substantiates evidence about the benefit of the reduction of required computational complexity and a reduction of implementation losses allowing for an increased throughput. This activity aims at designing, implementing and testing a demodulator software prototype based on ANNs for performing physical layer tasks: demodulation, de-mapping and channel decoding. Some of the main tasks to be performed are the design and implementation of the ANNs, acquisition of collected/generated data, data cleaning and pre-processing, topology definition, identification of learning algorithm, ANN training, evaluation and validation of the resulting model.