Demodulation of Sparse PPM Signals with Low Samples Using Trained RIP Matrix
This work addresses the challenge of efficient signal demodulation in communication systems, offering a domain-specific incremental improvement for sparse PPM signals.
The paper tackled the problem of demodulating sparse pulse position modulation (PPM) signals with fewer samples by training a neural network using the restricted isometry property (RIP) bound as a cost function to optimize measurements for classification. The results showed that the proposed method outperforms random measurements and requires less samples than the optimum matched filter demodulator, with some performance loss, and eliminates the need for an equalizer in multipath channels.
Compressed sensing (CS) theory considers the restricted isometry property (RIP) as a sufficient condition for measurement matrix which guarantees the recovery of any sparse signal from its compressed measurements. The RIP condition also preserves enough information for classification of sparse symbols, even with fewer measurements. In this work, we utilize RIP bound as the cost function for training a simple neural network in order to exploit the near optimal measurements or equivalently near optimal features for classification of a known set of sparse symbols. As an example, we consider demodulation of pulse position modulation (PPM) signals. The results indicate that the proposed method has much better performance than the random measurements and requires less samples than the optimum matched filter demodulator, at the expense of some performance loss. Further, the proposed approach does not need equalizer for multipath channels in contrast to the conventional receiver.