Energy-Efficient Spectral Analysis Method Using Autoregressive Model-Based Approach for Internet of Things

Abstract

This paper presents an energy-efficient spectral analysis method for the Internet of things. The objective of this work is to reduce the energy consumption of edge devices. The proposed method uses an autoregressive (AR) model for spectral analysis instead of the discrete Fourier transform, and its calculation process is distributed to the edge device and a base station by considering the energy consumption tradeoff of the data processing and the data communication. In this work, the Yule-Walker method is employed for the AR coefficient calculation. The calculation process of Yule-Walker method can be divided into two parts: an autocorrelation calculation and an AR coefficient calculation. The autocorrelation calculation is implemented in the edge devices, and its dedicated hardware is designed using Verilog HDL. Meanwhile, the AR coefficient is calculated in the base station and is used for the spectral analysis. According to this distributed processing approach, the energy consumption of the edge device can be reduced compared with conventional DFT approaches using the fast Fourier transform (FFT). The system level energy consumption is evaluated assuming the IoT edge device, which has a wireless transceiver using Bluetooth low energy. The evaluation results show that the proposed method can reduce 79% of the edge device energy consumption for spectral analysis in a practical application. Index Terms-Autoregressive (AR) model, data compression, edge computing, energy efficient devices, low power devices and circuits, spectral analysis, Yule-Walker method

Spectral estimator effects on accuracy of speed-over-ground radar

“…Furthermore, high model order increases the variance of the spectrum [23]. A suitable p-value that can sufficiently exhibit the properties of the signal but is not too large can be determined using the Akaike information criteria (AIC) and it is given by (10):…”

“…Spectral estimation based on the autoregressive (AR) model has also been evaluated for SoG radar. The choice of using the AR method is largely motivated by improved frequency resolution for short data length [10]. Unlike the FFT approach, this method models the data outside the processing length, which improves the spectral resolution of the observed signal [11].…”

A Performance Evaluation of Periodic Signal Analysis by ARS Compared with Frequency Analysis by FFT

“…Under low SNR conditions, ARS requires large β. Table 2 shows the relationship between α, the degree of FFT, andn F , the frequency index corresponding to the period resolution 0.01 s, as defined in (16). Additionally, the values ofn F R F Fig.…”

“…In such systems, the battery life of the sensor node is crucial in practice [15]. Therefore, the analysis should not be complicated in order to suppress the power consumption [16]. Although FFT is known as a simple method for frequency analysis [17], another simpler approach for signal analysis must be able to contribute to the edge computing.…”

ACOR: On the Design of Energy-Efficient Autocorrelation for Emerging Edge Applications

“…The ACF is commonly employed in various algorithms, such as autoregressive modeling, power spectral analysis (PSA) , linear predictive coding (LPC), and signal synchronization which are pivotal for commonly used signal processing applications, like bio-signal analysis [1][2][3], telecommunications [4], lidar systems [5] and audio processing/compression algorithms [6,7]. While providing key insights, the ACF calculation is considered to be one of the foremost energy-consuming parts of these algorithms.…”