Enhancing circuit adaptability in VLSI using hybrid optimisation for functional unit selection and resource allocation in high-level synthesis
M. Thillai Rani, K Pradeep, Ramah Sivakumar, S. Suresh Kumar · International Journal of Bio-Inspired Computation · 2025
Very-large-scale integration (VLSI) plays a crucial role in integrating transistors into a single chip, yet variations in process-voltage-temperature (PVT) present challenges for accuracy. In this manuscript, enhancing circuit adaptability in VLSI using hybrid optimisation for functional unit selection and resource allocation in high-level synthesis (VEA-Hyb-TSGTO-HLS) is proposed. It begins with data flow analysis (DFA), where the behavioural inputs of the VLSI circuits are analysed to gather crucial information about data flow and operation dependencies. Then volcano eruption algorithm (VEA) is used to determine the optimal functional unit based on variations in PVT, ensuring adaptability to changing conditions. Finally, hybrid transient search and group teaching optimisation algorithm (Hyb-TSGTO) is used to estimate the resource allocation. The proposed VEA-Hyb-TSGTO-HLS approach has achieved 24.6%, 21.4%, and 14.5% lesser cost and 15.6%, 18.8%, and 19.3% higher FU selection accuracy for using s38584 circuit when compared with the existing state of art methods.