Towards Efficient Channel Reconstruction with Spiking Neural Networks for MIMO Systems

In this research internship, I implemented spiking neural networks (SNNs) for the compression of Channel Impulse Response (CIR) signals in wireless communication systems. The primary objective was to reduce bandwidth usage by efficiently encoding CIR data, which is particularly relevant given the increasing data load in modern wireless networks. After compression, the CIR was transformed into the Channel Frequency Response (CFR) within a Multiple-Input Multiple-Output (MIMO) system. This approach aims to support more efficient data transmission by alleviating bandwidth constraints through biologically inspired, event-driven neural computation.