Edge computing for resilient and low-latency smart meter data processing

Abstract

With the more and more application of smart meters in modern energy systems, the processing requirements for large-scale real-time data have increased significantly. Traditional cloud-centric architectures could be influenced by network bandwidth and transmission latency in practical applications, leading to response delays and service interruptions. This thesis proposes an edge computing framework for robust and low-latency data processing in smart meters to support efficient energy management. This framework combines IoT-enabled smart meters with edge devices, performing data preprocessing and real-time analysis at the edge, and transmitting critical data to the cloud platform for monitoring and controlling.

By offloading time-sensitive tasks closer to the data source, the constructed framework aims to reduce end-to-end processing latency and improve service quality in the event of network fluctuations or outages. This thesis evaluates the framework, focusing on metrics such as processing latency, data availability and data transmission reliability. The result shows that, compared to a completely cloud-based processing model, the edge computing-based approach offers obvious improvements in response speed and system robustness.