Author: Amit Bansode, Neha Kulkarni, Kunal Jadhav
Abstract:The exponential growth of data generated by Internet of Things (IoT) devices has exposed the limitations of cloud-centric computing models, particularly in terms of latency, bandwidth consumption, privacy, and reliability. Edge computing has emerged as a promising paradigm that brings computation closer to data sources, enabling real-time analytics and intelligent decision-making. Embedded systems form the core of edge nodes, providing computing, storage, and communication capabilities under strict power and resource constraints. This paper presents a detailed study of embedded system architectures designed for highperformance edge computing. It discusses hardware accelerators, heterogeneous processing platforms, memory hierarchies, and software frameworks that enable low-latency and highthroughput edge applications. A reference architecture is proposed and evaluated using a realtime video analytics use case. Experimental analysis demonstrates that optimized embedded edge architectures significantly reduce latency and network load compared to cloud-only approaches.
Keywords: Edge computing, Embedded systems, Heterogeneous architecture, Hardware accelerators, IoT
Full Issue
| View or download the full issue | PDF 82-87 |