Low-power activity monitoring
AssigneeBiobit, Inc.
InventorsRohit Mittal, Neeraj Jhanji
Life of this patent
Estimating the term…
Abstract
Some examples include a secure low-power sensor platform able to be used for activity monitoring or other purposes. For instance, a sensor system may sense sparse signals using compressed sensing. As one example, compressed sensing may be used to obtain sparse signals from analog sensor signals received from one or more sensors. The data obtained by compressed sensing may be sent subsequently to a computing device to reconstruct the sensor signal. Examples herein enable lower power usage, lower cost, and lower weight than conventional devices, while also enabling processing advantages, such as less overall data to process and lower data storage utilization.
Claims22
A sensor block comprising:
- an amplifier to receive a sensor signal from a sensor and output an amplified signal;
- a compressed sensing analog-to-digital conversion block coupled to the amplifier, the compressed sensing analog-to-digital conversion block including a random number generator that generates random values to enable an analog-to-digital converter (ADC) to take random samples of the amplified signal based on the random values, and output compressed sensing data comprising the random samples;
- a post processor coupled to the compressed sensing analog-to-digital conversion block to receive the compressed sensing data and remove noise from the compressed sensing data by active thresholding;
- and
- a communication interface coupled to the post processor to receive the compressed sensing data from the post processor and send the compressed sensing data to a computing device, wherein the sensor block further utilizes the random number generator for encrypting the compressed sensing data prior to sending to the computing device.
A system for collecting and processing sensor data, the system comprising:
- a sensor block comprising:
- a compressed sensing analog-to-digital conversion block to receive a sensor signal, the compressed sensing analog-to-digital conversion block including a random number generator that causes a circuit to take random samples of the sensor signal based on randomly generated values from the random number generator and output compressed sensing data comprising the random samples;
- a post processor coupled to the compressed sensing analog-to-digital conversion block to receive the compressed sensing data and remove noise from the compressed sensing data by active thresholding;
- and
- a communication interface coupled to the post processor to receive the compressed sensing data from the post processor and send the compressed sensing data to a computing device, wherein the sensor block further utilizes the random number generator for encrypting the compressed sensing data prior to sending to the computing device.
A method of collecting and processing sensor data, the method comprising:
- obtaining a sensor signal from a sensor carried by a person;
- generating compressed sensing data using a compressed sensing analog-to-digital conversion block that uses values generated by a random number generator for taking random samples of the sensor signal at a frequency below a Nyquist rate;
- processing the compressed sensing data using a post processor to remove noise from the compressed sensing data by active thresholding, utilize the random number generator to encrypt the compressed sensing data, and packetize the encrypted compressed sensing data;
- and
- sending the packetized encrypted compressed sensing data to a computing device over a wireless connection.
Dependencies are inferred from the claim text.