Compressive multiplexing for radar data

A radar data processing system, comprising:

• a memory configured to store radar data in a data cube having bins distributed in positions described by indices along dimensions including a slow-time dimension, a fast-time dimension, and a channel dimension;
• and
• coding and combining circuitry configured to:
• divide the data cube along a first dimension into one or more zones;
• for each zone,
• select a number of data matrices, each data matrix comprising data vectors having bins that share an index value in a second dimension, wherein the number is based on a compression factor, further wherein the number of data matrices selected for a first zone is different from the number of data matrices selected for a second zone;
• form sets of data matrices containing the number of data matrices;
• for each set of data matrices, for each data matrix, apply a signal processing code to data vectors in the data matrix to generate a coded data matrix, wherein a signal processing coding scheme for data vectors in a data matrix is the same and a signal processing coding scheme for different data matrices is different;
• and
• combine the coded data matrices in the set in the second dimension to generate a compressed data matrix for the zone;
• and
• store the compressed data matrices for the one or more zones in the memory, wherein the compressed data matrices correspond to a compressed data cube.

Coding and combining circuitry configured to:

• read a first data vector and a second data vector stored in a data cube;
• apply a first signal processing coding scheme to a first data vector to generate a first coded data vector;
• apply a second signal processing coding scheme, different from the first signal processing coding scheme, to the second data vector to generate a second coded data vector;
• combine the first coded data vector and the second coded data vector to generate a compressed data vector;
• and
• store the compressed data vector.