Software Package for Compessing Big Data from BSA Radio-telescope

Astronomy is one of the fundamental sciences that studies celestial objects from publicly known such as planets, stars, comets to low studied phenomena including fast radio burst (FRB) events and pulsar objects. In Pushchino radio astronomy observatory the radio-telescope BSA is located, that registers radiation from the sky providing about 1 Mb of data per second or about 30 Tb of data per year. The observatory does not have big computing resources like clusters or super-computers to process such amount of data, and in result, by the 2019 year there has been collected more than 130 Tb of data, but it has not been processed. This work includes developing the software to process 130 Tb of data by calculating particular metrics on small data segments (for example, to calculate minimum, maximum, average and variance on each segment that represents 10 seconds of registered data), that in result will provide scientists with 10-100 times less amount of data which could be processed by ordinal personal computers. The software also was developed to be able to work on ordinal personal computers and to raise the performance there was used OpenCL technology, which also allowed to process data on both CPU and GPU devices. The result of the executed program is a file containing calculated metrics, that will be analyzed by observatory workers in order to find interesting objects such as pulsars. To visualize output data there was developed an additional program that allows drawing charts based on output data.