A Systematic Methodology for Time-multiplexing Algorithms on a Reconfigurable System-on-Chip

This paper introduces a methodology for hardware-software co-implementation, designed to simplify the System-on-Chip (SoC) design process when attempting to time-multiplex hardware. The approach streamlines the transformation of complex algorithms into efficient SoC implementations by providing a recipe for seamless hardware-software partitioning. As a case study, the methodology is applied to the Extrapolated Single Propagation Particle Filter algorithm for geo-locating sources of radio frequency interference and results demonstrate a considerable speed-up when using a time-multiplexed implementation.