Analysis of different multiprocessor architectures for radar signal processing with performance metrics

Radar signal processing requires platform with high number crunching capability for a real time performance. Towards this, multiprocessor based architectures plays a significant role in achieving real time processing speed for voluminous radar data. Also, different radar signal processing algorithms are amenable to partition in either data or algorithm domain to achieve execution speedup. In this paper, we intend to analyze different multiprocessor based architectures quantitatively for performance metrics and analytical relationship between performance metrics like load imbalance and efficiency of multiprocessor platforms are also brought out. Depending upon particular radar application, these performance metrics are taken into account before deciding the architecture. The case study for different architectures for their efficiency is also carried out.