Achieving Low Complexity Neural Decoders via Iterative Pruning
This work addresses the problem of computational efficiency for neural decoders on edge devices, but it is incremental as it builds on existing pruning methods.
The paper tackles the high complexity of neural decoders in communications by applying iterative pruning to reduce weight count, achieving lower latency and complexity while maintaining accuracy, and introduces semi-soft decision decoding to improve bit error rate performance.
The advancement of deep learning has led to the development of neural decoders for low latency communications. However, neural decoders can be very complex which can lead to increased computation and latency. We consider iterative pruning approaches (such as the lottery ticket hypothesis algorithm) to prune weights in neural decoders. Decoders with fewer number of weights can have lower latency and lower complexity while retaining the accuracy of the original model. This will make neural decoders more suitable for mobile and other edge devices with limited computational power. We also propose semi-soft decision decoding for neural decoders which can be used to improve the bit error rate performance of the pruned network.