Compression of Visibility Data for Murchison Widefield Array

TL;DR

This paper introduces a simple binary compression method for MWA visibility data, significantly reducing storage needs by up to 3.1 times, thereby addressing operational costs.

Contribution

A novel binary compression technique tailored for MWA visibility data, improving storage efficiency with minimal complexity.

Findings

Compression ratio up to 1:3.1

Reduces data storage costs

Simple implementation

Abstract

The Murchison Widefield Array (MWA) is a new low frequency radio telescope operating on the Square Kilometre Array site in Western Australia. MWA is generating tens of terabytes of data daily. The size of the required data storage has become a significant operational limitation and cost. We present a simple binary compression technique and a system for the floating point visibility data developed MWA. We present the statistics of the impact of such compression on the data with the typical compression ratio up to 1:3.1.