An Information Centric Framework for Weather Sensing Data

TL;DR

This paper proposes an information-centric framework using Named Data Networking to improve data retrieval and weather prediction accuracy from distributed radars, addressing connectivity issues in remote locations.

Contribution

It introduces a hierarchical naming scheme for weather data in NDN, enhancing data access, quality, and prediction compared to traditional TCP/IP mechanisms.

Findings

NDN improves data quality over TCP/IP

Hierarchical naming enables effective data retrieval

Enhanced weather prediction accuracy

Abstract

Weather sensing and forecasting has become increasingly accurate in the last decade thanks to high-resolution radars, efficient computational algorithms, and high-performance computing facilities. Through a distributed and federated network of radars, scientists can make high-resolution observations of the weather conditions on a scale that benefits public safety, commerce, transportation, and other fields. While weather radars are critical infrastructure, they are often located in remote areas with poor network connectivity. Data retrieved from these radars are often delayed or lost, or even lack proper synchronization, resulting in sub-optimal weather prediction. This work applies Named Data Networking (NDN) to a federation of weather sensing radars for efficient content addressing and retrieval. We identify weather data based on a hierarchical naming scheme that allows us to explicitly access desired files. We demonstrate that compared to the window-based mechanism in TCP/IP, an NDN based mechanism improves data quality, reduces uncertainty, and enhances weather prediction. Our evaluation demonstrates that this naming scheme enables effective data retrieval, while compared to the window-based mechanism in TCP/IP, an NDN based mechanism improves data quality, reduces uncertainty, and enhances weather prediction.