Accelerating and Improving Simulation Performance in Communication Systems Modelling through Parallel Computing and Clustering
Existing 5G communication systems suffer from two major problems: the need for better spectrum efficiency and lesser adjacent channel interference. Thus, development of novel waveform techniques to overcome these problems is a major topic of research amongst scholars and it requires carrying out Monte Carlo simulations in MATLAB© (by MathWorks) to measure the Bit Error Rate (BER) of these communication models. As most of these simulations require millions of computations, they take a significantly long time to run (for example, days) as they run on single-core machines and carry out the computations serially. The main objective of this research is to reimplement current scripts using various parallel computing techniques in MATLAB to study which one is the best suited for this particular type of simulations while also scaling these scripts onto a multi-core cluster to further improve the execution time.