Rapid computation and analysis using a multiple core to enhance SSVEP based system

Brain-computer interface (BCI) technology is communication system is rely a pathway to explore the brain activities to external world. The BCI technique make possible to monitor some physical processes that occur within the brain activity that correspond to certain forms of flickering light through stimuli. Observed a thousand of brain activities are firing instantaneous which allowed the BCI system to explicit one or more signals are controlled on computer command or dominance any other devices. Electroencephalogram (EEG) signal based BCI facing a serious challenge in online applications which are slower reaction and consume more time at computational analysis and extraction. Prosperous advance techniques based on EEG raw-data to extract feature with distributed computing system which are offered a promises result that overcomes these gaps. Therefore, have been developed a BCI prototype that realized by reliable capability which take decision in real time or predicted an inclination in real life application. In previous studies were employed a single CPU system, which is revealed decent performance for smaller dataset; in other hand the open multi-processing (OpenMP) platform provide a high performance computing in more accuracy and supplemental precisely outcome within a large datasets. The main concept of parallelize computing that can be separate the tasks individually which is allowed to parallelized process based on multiple cores. In this, work conclude two approaches which are utilized a high performance computing (HPC) to realize a faster analysis reaction of brain activities and recognition based on evoked SSVEP signal by exploring the Hilbert transform (HT) and quadrature amplitude demodulation (QAD) techniques depend on patterns detection; however have been employed a five frequencies to extract short-term Fourier transform (STFT) feature based on four type filters using windowing function. Both approaches were adapted into HPC technique to distinguish the extraction and execution time.