- Low power ASIC - ASIC stands for "Application Specific Integrated Circuit". A low power ASIC is designed for specific applications with low power consumptions. A Body Sensor Network (BSN) ASIC usually contains both digital and analogue units in a single chip.
- AFE - Analogue Front-End (AFE) is the part of the ASIC that interfaces directly with sensors. It normally contains amplifiers, filters, comparators, Analogue-to-Digital converters, voltage sources, and current sources. An AFE amplifies bio-signals from the body, filters out unwanted noises and signals, and then converts signals to digital forms for data processing.
- On-Node signal processing - On-Node processing is a method used in wireless sensing that analyses data directly on the sensor node before it is transmitted to other parts of the network. As such, data abstraction is done at the node level and this minimises wireless data transfer (which is usually the most power consuming part of the network).
“Use body as the medium, a source of inspiration, and energy to provide long-term, continuous sensing, monitoring and intervention – GZ Yang.” Body Sensor Network (BSN) was coined to harness several allied technologies that underpin the development of pervasive sensing for healthcare, wellbeing, sports and other applications that require “ubiquitous” and “pervasive” monitoring of physical, physiological, and biochemical parameters in any environment without activity restriction. Key to the development of BSNs are technologies that address miniaturised biosensor design suitable for both wearable and implantable devices, biocompatibility and materials to ensure long-term deployment, low-power wireless communication, integrated circuits and systems, power scavenging techniques from the body, autonomous sensing, and standards and integration.
A BSN ASIC typically includes multiple AFEs to interface with transducers (sensors), to permit analogue/mixed-signal processing to reduce data dimensionality, and to provide a microprocessor for reconfigurable control, data processing and communication. The primary mode of operation of the AFEs will be to fix a voltage across the transducer and measure the flow of current through it. An example transducer is a device that has a resistance proportional to the pressure applied to the device. The output signal of an AFE will then be passed to an analogue/mixed-signal circuitry that performs signal conditioning and processing (both linear and non-linear) as required by the sensing application. This processing is likely to be a combination of both continuous-time, discrete-time and event-based methods. The output signal is then quantised and processed by a digital microprocessor.
At the core of a BSN ASIC is a microcontroller. This collects data from the embedded analogue system, provides timing and control signals to the embedded analogue system, interfaces to external sensors, manages the power consumption of the system, and controls the communication with other sensing nodes.
The ASIC shown on the left (implemented in TSMC 0.18µm technology) presents an analogue intelligent front-end integrated with an ultra-low power ARM cortex M0 CPU (Central Process Unit) with 2KB SRAM and 32KB OTP (One Time Programmable) flash memory.
This chip was designed to provide a multi-functional AFE based on current conveyor topology that can be used for continuous monitoring of physiological signals such as PPG (Photoplenthysmography) and ECG (Electrocardiogram). It is designed with a 3-bit programmable threshold voltage peak detector, which works in the micro-amp region and provides early signal detection before waking up the entire digital system. It incorporates an AFE power control module to minimise the power consumption of the chip.
What are we using it for?
We are using the ASIC design combined with sensors to measure heart rate, tissue oxygenation, Photoplenthysmography (PPG) and biomotion. The chip is designed, however, for generic biosensing applications.