This paper presents a comparison of various low-power techniques applied to filters, specifically focusing on low-power linear phase finite impulse response (FIR) filters designed for system-on-chip (SoC) applications. The study comprehensively evaluates the impact of dynamic voltage scaling (DVS), power gating (PG), and clock gating (CG) on various FIR filter architectures, highlighting significant power savings and performance trade-offs. Low-power design strategies must be used for efficient signal processing in power-sensitive applications. Three digital filters are compared in this study: a moving average filter, a 6-tap low-pass finite impulse response (FIR) filter, and a 15-tap low-pass FIR filter. The main objective is to assess how low-power methods like dynamic voltage scaling, power gating, and clock gating affect these filters. Using Verilog HDL and the Xilinx Vivado design suite, a design and simulation approach was employed to assess the effectiveness of these power-saving techniques. After synthesis, the performance and power consumption metrics were taken out to evaluate how each technique affected the different filter topologies. Our results show the many levels of power savings that can be achieved and offer a smart analysis of the trade-offs between lower power consumption and signal processing integrity, including filter response alterations and latency introduction. This work offers significant insights for system designers by demonstrating how to take advantage of the features of the Xilinx Vivado toolset and Verilog coding to maximize low-power strategies for particular filter designs, which are essential for portable electronic devices and power-efficient embedded systems.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Exploring Low-Power Digital Filter Design Techniques for System-On-Chip (SoC) Applications

  • Sameer Singh,
  • Pallabi Sarkar,
  • Ankur Beohar

摘要

This paper presents a comparison of various low-power techniques applied to filters, specifically focusing on low-power linear phase finite impulse response (FIR) filters designed for system-on-chip (SoC) applications. The study comprehensively evaluates the impact of dynamic voltage scaling (DVS), power gating (PG), and clock gating (CG) on various FIR filter architectures, highlighting significant power savings and performance trade-offs. Low-power design strategies must be used for efficient signal processing in power-sensitive applications. Three digital filters are compared in this study: a moving average filter, a 6-tap low-pass finite impulse response (FIR) filter, and a 15-tap low-pass FIR filter. The main objective is to assess how low-power methods like dynamic voltage scaling, power gating, and clock gating affect these filters. Using Verilog HDL and the Xilinx Vivado design suite, a design and simulation approach was employed to assess the effectiveness of these power-saving techniques. After synthesis, the performance and power consumption metrics were taken out to evaluate how each technique affected the different filter topologies. Our results show the many levels of power savings that can be achieved and offer a smart analysis of the trade-offs between lower power consumption and signal processing integrity, including filter response alterations and latency introduction. This work offers significant insights for system designers by demonstrating how to take advantage of the features of the Xilinx Vivado toolset and Verilog coding to maximize low-power strategies for particular filter designs, which are essential for portable electronic devices and power-efficient embedded systems.