Summary of Neuromorphic Wireless Split Computing with Multi-level Spikes, by Dengyu Wu et al.
Neuromorphic Wireless Split Computing with Multi-Level Spikes
by Dengyu Wu, Jiechen Chen, Bipin Rajendran, H. Vincent Poor, Osvaldo Simeone
First submitted to arxiv on: 7 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Information Theory (cs.IT); Neural and Evolutionary Computing (cs.NE)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper explores neuromorphic computing, inspired by biological processes, which leverages spiking neural networks (SNNs) to perform inference tasks with significant efficiency gains for sequential data workloads. By embedding a small payload within each spike exchanged between neurons, recent advances in hardware and software have shown enhanced accuracy without increased energy consumption. To scale neuromorphic computing, split computing is proposed, where an SNN is partitioned across two devices, requiring information transmission about spikes generated by output neurons. This establishes a trade-off between multi-level spikes carrying additional payload information and communication resources required for transmitting extra bits. The paper presents the first comprehensive study of a neuromorphic wireless split computing architecture employing multi-level SNNs. Digital and analog modulation schemes are proposed for an orthogonal frequency division multiplexing (OFDM) radio interface to enable efficient communication. Simulation and experimental results using software-defined radios reveal performance improvements achieved by multi-level SNN models, providing insights into the optimal payload size as a function of connection quality between transmitter and receiver. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper is about using computers that work like our brains to do tasks more efficiently. It uses special networks called spiking neural networks (SNNs) that are inspired by how our brain works. These SNNs can be split across two devices, which helps with communication and accuracy. The paper shows how this works and what kind of results you get when using it. |
Keywords
* Artificial intelligence * Embedding * Inference
