Compressive sensing systems and methods using edge nodes of distributed computing networks

1. A method for compressive sensing using an edge node of a resource-constrained distributed computing network, wherein the edge node is provided with a fixed network capacity (Ø) and is configured to execute the method comprising:

collecting a raw data signal (X) continuously by a sensor of the edge node of the resource-constrained distributed computing network;

dynamically updating a signal energy indicator (ε) that quantifies a variance in the raw data signal;

determining one or more compression characteristics as a function of the signal energy indicator as the signal energy indicator is updated;

subsampling the raw data signal in accordance with current values of the one or more compression characteristics to create a compressed data signal;

comparing a size of the compressed data signal to the fixed network capacity (Ø) and further subsampling the compressed data signal to further compress the compressed data signal if the size of the compressed data signal is greater than a size permitted by the fixed network capacity; and

transmitting an output that includes the compressed data signal to a centralized node of the resource-constrained distributed computing network.

2. The method of claim 1 , wherein the one or more compression characteristics includes a sampling frequency (δ), a signal window length (N), or a combination including at least one of the foregoing.

3. The method of claim 2 , wherein the signal window length is also determined based on the fixed network capacity (Ø) of the distributed computing network available to the edge node.

4. The method of claim 2 , wherein the subsampling includes at least one of randomly subsampling and randomly down-sampling the raw data signal with respect to the sampling frequency.

5. The method of claim 4 , further comprising generating random numbers with a random number generator and utilizing the random numbers as indices for the at least one of subsampling and down-sampling.

6. The method of claim 5 , wherein the random number generator is a seedable random number generator, such that the random numbers can be recreated by seeding a second instance of the random number generator at a centralized node during decompression of the compressed data signal.

7. The method of claim 6 , further comprising sending additional samples or packets from the raw data signal if the size of the compressed data signal is less than the size permitted by the fixed network capacity.

8. The method of claim 1 , further comprising decompressing the compressed data signal by the centralized node.

9. The method of claim 8 , further comprising estimating energy in the raw data signal based on a size of the compressed data signal with respect to the raw data signal.

10. The method of claim 9 , further comprising selecting a decompression algorithm based on the energy estimated in the raw data signal.

11. An edge node for a resource-constrained distributed computing network, wherein the edge node is provided with a fixed network capacity (Ø), the edge node comprising:

a communication module configured to enable data communication over the resource-constrained distributed computing network;

a sensor configured to continuously collect a raw data signal (X) related to one or more parameters of an environment local to the edge node; and

a controller configured to:

receive the raw data signal;

dynamically update a signal energy indicator (ε) that quantifies a variance in the raw data signal;

determine one or more compression characteristics as a function of the signal energy indicator as the signal energy indicator is updated;

subsample the raw data signal in accordance with current values of the one or more compression characteristics to create a compressed data signal;

compare a size of the compressed data signal to the fixed network capacity (Ø) and further subsample the compressed data signal to further compress the compressed data signal if the size of the compressed data signal is greater than a size permitted by the fixed network capacity; and

transmit an output that includes the compressed data signal to a centralized node on the resource-constrained distributed computing network.

12. The edge node of claim 11 , wherein the edge node comprises a luminaire.

13. The edge node of claim 12 , wherein the sensor is a motion detection sensor.

14. A system including the edge node of claim 11 and a centralized node, wherein the centralized node is a gateway, a server, a cloud computing implementation, or a combination including at least one of the foregoing.