Photonic computing to accelerate data processing in wireless communications

• DOI: 10.1364/oe.423747
• Publication Date: 2021-06-21T15:00:06Z
• Authors: Mahsa Salmani; Armaghan Eshaghi; Enxiao Luan; Sreenil Saha
• Keywords: FOS: Computer and information sciences; Emerging Technologies (cs.ET); Computer Science - Information Theory; Information Theory (cs.IT); 0202 electrical engineering, electronic engineering, information engineering; Computer Science - Emerging Technologies; 02 engineering and technology; 7. Clean energy

Massive multiple-input multiple-output (MIMO) systems are considered as one of the leading technologies employed in next generations wireless communication networks (5G), which promise to provide higher spectral efficiency, lower latency, and more reliability. Due massive number devices served by base stations (BS) equipped with large antenna arrays, massive-MIMO need perform high-dimensional signal processing a considerably short amount time. The computational complexity such data processing, while satisfying energy latency requirements, is beyond capabilities conventional widely-used digital electronics-based computing, i.e., Field-Programmable Gate Arrays (FPGAs) Application-Specific Integrated Circuits (ASICs). In this paper, speed lossless propagation light exploited introduce photonic computing approach that addresses high required systems. proposed based on implementation multiply accumulate (MAC) operation achieved broadcast-and-weight (B&W) architecture. B&W protocol limited real positive values MAC operations. work, preprocessing steps developed enable architecture accept any arbitrary input. This requirement for typically deal complex values. Numerical analysis shows performance system not degraded architecture, it provides significant improvements time efficiency compared most powerful Graphics Processing Units (GPUs).