Edge-Enabled Internet Weather Mapping for IoT Traffic Visualization

Abstract

The Internet of Things (IoT) has grown at a very fast rate resulting in a very high heterogeneous network traffic, which causes the dynamism in the edge networks. These swings have a huge impact to the end-to-end quality service and the health of the entire Internet. This paper suggests an edge based Internet weather mapping architecture that can be used to visualise and interpret IoT traffic patterns in real time. The suggested system combines edge-based passive flow tracking and the cloud-based visualisation pipeline that links the key performance indicators, including latency, throughput, and packet loss. Through temporal and spatial changes in IoT data streams, the system produces visualised weather maps which can represent congestion hotspots, transmission issues, and degradation of connectivity in a given region. The framework makes use of the adaptive edge analytics to minimise the latency involved in gathering the data and also utilises self-organising map (SOM)-based clustering to improve cognitive readability of the performance dynamics. Experimental deployment results in the various IoT zones reveal enhanced visualisation granularity and predictive awareness which help operators of the network to optimise the network proactively. The architecture fills the divide between the localized IoT data and the visualization of the global network, and provides a scalable and understandable model of future internet telemetry.