DNA Cryptography-Based Secured Weather Prediction Model in High-Performance Computing

Abstract

This paper discusses the design of a DNA cryptography-based secured weather prediction model by the use of supercomputing or cluster type computing environment. The model is based on Markov’s chain. The supercomputer clusters are mainly required to run high-resource and time-demanding applications which a single computer cannot run. Use of supercomputers ensures faster and efficient computational power. High-performance computing (HPC) can be used to build a centralized file server for the Web and can easily process the information with its high processing speeds. A weather prediction system generally involves a large amount of past data to be processed over for an efficient prediction of the future weather. This paper lays emphasis on the use of Markov’s chain to develop a weather prediction model which depends on a larger set of input data types and can be implemented on a HPC system environment for a lesser computational time and higher accuracy. A flexible algorithm is proposed for weather prediction named averaged transit prediction (ATP) algorithm here. This model has been further integrated with a novel DNA cryptography-based algorithm named decimal bond DNA (DBD) algorithm for secured transmission of data between different processors of HPC. The simulated results on test bed formed by connecting five nodes in parallel mode forming supercomputing environment and having a performance of 0.1 Tflops gave predicted temperature, humidity, and wind speed for three different days with an accuracy of 85–95%.