Design Space Exploration of DSP Hardware Using Adaptive PSO and Bacterial Foraging for Power/Area-Delay Trade-Off

Abstract

Digital signal processing (DSP) hardware is ubiquitous in the current generation of consumer electronics systems including camera, camcorders, set-top boxes, smartphones, etc. The very large-scale integration (VLSI) design process of DSP hardware is entirely dependent on high-level synthesis framework that comprises design space exploration (DSE) of power/area-delay trade-off. Since DSP hardware is application specific by nature, and thus, exploration of its low power, high performance architectural solution is crucial. However, the exploration process is intricate and involves a number of convoluted factors such as modelling of the objective function (such as power and delay), accuracy/efficiency of the optimization framework, loop dependency, data pipelining, seed encoding, scheduling algorithm, resource binding, ability to escape local minima and terminating criteria. In this chapter, we present a number of emerging evolutionary design space exploration techniques based on bacterial foraging and particle swarm optimization algorithm that is capable to consider the aforesaid complex factors while performing power/area—delay trade-off of DSP hardware. The chapter also discusses the analysis on case studies for each DSE technique with respect to power–delay trade-off. © Springer Nature Singapore Pte Ltd 2020.