Abstract:Dynamic voltage and frequency scaling (DVFS) is a technique used to optimize energy consumption in ultra-low-power embedded systems. To ensure sufficient computational capacity, the system must scale up its performance settings. The objective is to conserve energy in times of reduced computational demand and/or when battery power is used. Fast Fourier Transform (FFT), Cyclic Redundancy Check 32 (CRC32), Secure Hash Algorithm 256 (SHA256), and Message-Digest Algorithm 5 (MD5) are focused functions that demand computational power to achieve energy-efficient performance. Selected operations are analyzed from the energy consumption perspective. In this manner, the energy required to perform a specific function is observed, thereby mitigating the influence of the instruction set or system architecture. For stable operating voltage scaling, an exponential model for voltage calculation is presented. Statistical significance tests are conducted to validate and support the findings. Results show that the proposed optimization technique reduces energy consumption for ultra-low-power applications from 27.74% to up to 47.74%.
engineering, electrical & electronic,computer science, information systems,physics, applied
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to reduce energy consumption in ultra - low - power embedded systems. Specifically, the paper explores how Dynamic Voltage and Frequency Scaling (DVFS) technology can optimize the energy efficiency of these systems. By reducing the operating frequency and voltage of the processor at low loads and increasing them at high loads, DVFS can significantly reduce energy consumption while ensuring computing power, thereby prolonging battery life and reducing heat generation.
### Main contributions of the paper:
1. **Energy analysis**: The paper conducted an energy analysis on computationally intensive operations such as FFT128, FFT32, CRC32, MD5 and SHA256 to evaluate the energy consumption under different loads.
2. **Dynamic voltage calculation model**: An exponential model was proposed to calculate the dynamic voltage value to ensure stable operation at different operating frequencies.
3. **Performance - power balance**: By dynamically adjusting the voltage and frequency, the optimal balance between performance and power consumption was achieved under different loads.
4. **Experimental verification**: The effectiveness of DVFS technology was verified through experiments, and the results showed that in ultra - low - power applications, the energy consumption was reduced by 27.74% to 47.74%.
### Key formulas:
- **Dynamic voltage calculation formula**:
\[
U_{\text{DVFS}}(f)=A\times e^{f / B}\quad\text{[V]}
\]
where \(A = 1.791011\) and \(B = 26.1804846\).
### Experimental setup:
- **Measuring equipment**: Keysight 34465A multimeter and GW INSTEK PSB - 1400L programmable DC power supply were used for measurement.
- **Microcontroller**: The ARM Cortex - M0+ ultra - low - power STM32L0 microcontroller was selected.
- **Development environment**: STM32CubeIDE was used for code generation and debugging.
### Results discussion:
- **Dynamic frequency adjustment (DFS)**: As the operating frequency increases, although the current increases, the overall energy consumption decreases because the execution time is significantly reduced.
- **Dynamic voltage and frequency adjustment (DVFS)**: After combining with dynamic voltage adjustment, the energy consumption is further reduced, especially in the case of low loads, the effect is more significant.
Through these studies, the paper provides an important reference for the design of ultra - low - power embedded systems, especially in application scenarios that require long - term operation and have limited energy.