The Future of Ultra-Low Power Signal Processing

One of my favorite quotes comes from the IEEE Signal Processing magazine in 2010. They attempted to answer the question: What does ultra-low power consumption mean? And they came to the conclusion that it is where the “power source lasts longer than the useful life of the product.”[1] It’s a great answer because it’s scalable. It applies equally to signal processing circuitry inside an embedded IoT device that can never be accessed or recharged and to signal processing inside a car where the petrol for the engine dominates the operating lifetime, not the signal processing power. It also describes exactly what a lot of science fiction has always envisioned: no changing or recharging of batteries, which people forget to do or never have enough batteries for. Rather, we have devices that simply always work.Figure 1

My research focuses on healthcare applications and creating “wearable algorithms”—that is, signal processing implementations that fit within the very small power budgets available in wearable devices. Historically, this focused on data reduction to save power. It’s well known that wireless data transmission is very power intensive. By using some power to reduce the amount of data that has to be sent, it’s possible to save lots of power in the wireless transmission stage and so to increase the overall battery lifetime.

This argument has been known for a long time. There are papers dating back to at least the 1990s based on it. It’s also readily achievable. Inevitably, it depends on the precise situation, but we showed in 2014 that the power consumption of a wireless sensor node could be brought down to the level of a node without a wireless transmitter (one that uses local flash memory) using easily available, easy-to-use, off-the-shelf-devices.[2]

This essay appears in Circuit Cellar 316, November 2016. Subscribe to Circuit Cellar to read project articles, essays, interviews, and tutorials every month!

Today, there are many additional benefits that are being enabled by the emerging use of ultra-low power signal processing embedded in the wearable itself, and these new applications are driving the research challenges: increased device functionality; minimized system latency; reliable, robust operation over unreliable wireless links; reduction in the amount of data to be analyzed offline; better quality recordings (e.g., with motion artifact removal to prevent signal saturations); new closed-loop recording—stimulation devices; and real-time data redaction for privacy, ensuring personal data never leaves the wearable.

It’s these last two that are the focus for my research now. They’re really important for enabling new “bioelectronic” medical devices which apply electrical stimulation as an alternative to classical pharmacological treatments. These “bioelectronics” will be fully data-driven, analyzing physiological measurements in real-time and using this to decide when to optimally trigger an intervention. Doing such as analysis on a wearable sensor node though requires ultra-low power signal processing that has all of the feature extraction and signal classification operating within a power budget of a few 100 µW or less.

To achieve this, most works do not use any specific software platform. Instead they achieve very low-power consumption by using only dedicated and highly customized hardware circuits. While there are many different approaches to realizing low-power fully custom electronics, for the hardware, the design trends are reasonably established: very low supply voltages, typically in the 0.5 to 1 V range; highly simplified circuit architectures, where a small reduction in processing accuracy leads to substantial power savings; and the use of extensive analogue processing in the very lowest power consumption circuits.[3]

Less well established are the signal processing functions for ultra-low power. Focusing on feature extractions, our 2015 review highlighted that the majority (more than half) of wearable algorithms created to date are based upon frequency information, with wavelet transforms being particularly popular.[4] This indicates a potential over-reliance on time–frequency decompositions as the best algorithmic starting points. It seems unlikely that time–frequency decompositions would provide the best, or even suitable, feature extraction across all signal types and all potential applications. There is a clear opportunity for creating wearable algorithms that are based on other feature extraction methods, such as the fractal dimension or Empirical Mode Decomposition.

Investigating this requires studying the three-way trade-off between algorithm performance (e.g., correct detections), algorithm cost (e.g., false detections), and power consumption. We know how to design signal processing algorithms, and we know how to design ultra-low power circuitry. However, combining the two opens many new degrees of freedom in the design space, and there are many opportunities and work to do in mapping feature extractions and classifiers into sub-1-V power supply dedicated hardware.


[1] G. Frantz, et al, “Ultra-low power signal processing,” IEEE Signal Processing Magazine, vol. 27, no. 2, 2010.
[2] S. A. Imtiaz, A. J. Casson, and E. Rodriguez-Villegas, “Compression in Wearable Sensor Nodes,” IEEE Transactions on Biomedical Engineering, vol. 61, no. 4, 2014.
[3] A. J. Casson, et al, “Wearable Algorithms,” in E. Sazonov and M. R. Neuman (eds.), Wearable Sensors, Elsevier, 2014.
[4] A. J. Casson, “Opportunities and Challenges for Ultra Low Power Signal Processing in Wearable Healthcare,” 23rd European Signal Processing Conference, Nice, 2015.


Alex Casson is a lecturer in the Sensing, Imaging, and Signal Processing Department at the University of Manchester. His research focuses on creating next-generation human body sensors, developing both the required hardware and software. Dr. Casson earned an undergraduate degree at the University of Oxford and a PhD from Imperial College London.

The Future of Biomedical Signal Analysis Technology

Biomedical signals obtained from the human body can be beneficial in a variety of scenarios in a healthcare setting. For example, physicians can use the noninvasive sensing, recording, and processing of a heart’s electrical activity in the form of electrocardiograms (ECGs) to help make informed decisions about a patient’s cardiovascular health. A typical biomedical signal acquisition system will consist of sensors, preamplifiers, filters, analog-to-digital conversion, processing and analysis using computers, and the visual display of the outputs. Given the digital nature of these signals, intelligent methods and computer algorithms can be developed for analysis of the signals. Such processing and analysis of signals might involve the removal of instrumentation noise, power line interference, and any artifacts that act as interference to the signal of interest. The analysis can be further enhanced into a computer-aided decision-making tool by incorporating digital signal processing methods and algorithms for feature extraction and pattern analysis. In many cases, the pattern analysis module is developed to reveal hidden parameters of clinical interest, and thereby improve the diagnostic and monitoring of clinical events.Figure1

The methods used for biomedical signal processing can be categorized into five generations. In the first generation, the techniques developed in the 1970s and 1980s were based on time-domain approaches for event analysis (e.g., using time-domain correlation approaches to detect arrhythmic events from ECGs). In the second generation, with the implementation of the Fast Fourier Transform (FFT) technique, many spectral domain approaches were developed to get a better representation of the biomedical signals for analysis. For example, the coherence analysis of the spectra of brain waves also known as electroencephalogram (EEG) signals have provided an enhanced understanding of certain neurological disorders, such as epilepsy. During the 1980s and 1990s, the third generation of techniques was developed to handle the time-varying dynamical behavior of biomedical signals (e.g., the characteristics of polysomnographic (PSG) signals recorded during sleep possess time-varying properties reflecting the subject’s different sleep stages). In these cases, Fourier-based techniques cannot be optimally used because by definition Fourier provides only the spectral information and doesn’t provide a time-varying representation of signals. Therefore, the third-generation algorithms were developed to process the biomedical signals to provide a time-varying representation, and   clinical events can be temporally localized for many practical applications.

This essay appears in Circuit Cellar 315, October 2016. Subscribe to Circuit Cellar to read project articles, essays, interviews, and tutorials every month!

These algorithms were essentially developed for speech signals for telecommunications applications, and they were adapted and modified for biomedical applications. The nearby figure illustrates an example of knee vibration signal obtained from two different knee joints, their spectra, and joint time-frequency representations. With the advancement in computing technologies, for the past 15 years, many algorithms have been developed for machine learning and building intelligent systems. Therefore, the fourth generation of biomedical signal analysis involved the automatic quantification, classification, and recognition of time-varying biomedical signals by using advanced signal-processing concepts from time-frequency theory, neural networks, and nonlinear theory.

During the last five years, we’ve witnessed advancements in sensor technologies, wireless technologies, and material science. The development of wearable and ingestible electronic sensors mark the fifth generation of biomedical signal analysis. And as the Internet of Things (IoT) framework develops further, new opportunities will open up in the healthcare domain. For instance, the continuous and long-term monitoring of biomedical signals will soon become a reality. In addition, Internet-connected health applications will impact healthcare delivery in many positive ways. For example, it will become increasingly effective and advantageous to monitor elderly and chronically ill patients in their homes rather than hospitals.

These technological innovations will provide great opportunities for engineers to design devices from a systems perspective by taking into account patient safety, low power requirements, interoperability, and performance requirements. It will also provide computer and data scientists with a huge amount of data with variable characteristics.

The future of biomedical signal analysis looks very promising. We can expect  innovative healthcare solutions that will improve everyone’s quality of life.

Sridhar (Sri) Krishnan earned a BE degree in Electronics and Communication Engineering at Anna University in Madras, India. He earned MSc and PhD degrees in Electrical and Computer Engineering at the University of Calgary. Sri is a Professor of Electrical and Computer Engineering at Ryerson University in Toronto, Ontario, Canada, and he holds the Canada Research Chair position in Biomedical Signal Analysis. Since July 2011, Sri has been an Associate Dean (Research and Development) for the Faculty of Engineering and Architectural Science. He is also the Founding Co-Director of the Institute for Biomedical Engineering, Science and Technology (iBEST). He is an Affiliate Scientist at the Keenan Research Centre at St. Michael’s Hospital in Toronto.