New XL MIMO system enables wave-domain neural network inference

Researchers have developed a novel approach to performing machine learning inference directly on wirelessly transmitted data by integrating neural network computations with the wireless channel itself. This method utilizes an eXtremely Large Multiple-Input Multiple-Output (XL MIMO) system equipped with Stacked Intelligent Metasurfaces (SIM) at the receiver. The system acts as a two-layer neural network, with the wireless channel forming the first layer and the SIM implementing the second, controllable layer. By leveraging the Extreme Learning Machine (ELM) framework, this architecture demonstrates universal approximation capabilities and achieves classification accuracy comparable to idealized digital models, showcasing the potential for low-complexity, wave-domain neural computations. AI

IMPACT This research could lead to more efficient and lower-complexity AI inference directly within wireless communication systems.

RANK_REASON This is a research paper detailing a new technical approach in signal processing and machine learning. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• extreme learning machine
• Kyriakos Stylianopoulos
• Metasurfaces for field control and detection
• Stacked Metasurfaces
• XL MIMO Extreme Learning Machines

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →