Digital beamforming arrays must transport digitized sample data to other locations for computation and the formation of final receive beams.
This can be accomplished by a number of different architectures. A central computer can take in all samples from every stream and form any number of digital beams, limited only by the amount of processing power available. But for very large arrays, a processor of this nature requires an enormous amount of digital input bandwidth and input connections.
Navy scientists have provided a solution to this bandwidth problem by the implementation of a three-threshold detection process for a digital phased array, leading to a reduction of digital I/O bandwidth requirements between subarray outputs and the beamformer. The system applies matched filtering and detection at each subarray which is possible with distributed architectures and does not impact digital beamforming.
In a radar example, preprocessing occurs at each node to obtain data on where targets of interest likely exist in a particular configuration of targets. Thus a smaller set of data is sent to a central processor that would be required of a traditional digital phased array radar system while retaining desired probability of detection and other pertinent performance parameters.
- By pulling some of the signal detection processing that occurs at the final processor in a pulse-detecting system into each node for pre-processing, an amount of digital sample data that must be sent to the final processor can be greatly reduced
- Matched filtering and threshold detection earlier in the beam former means that signal to noise ratio (SNR) is much lower than typical for radar systems
- A three-threshold detection process enables detection sensitivity comparable to standard coherent radar
- Schema can be applied across thousands of digital nodes in a phased array
- US application number 20170324440 available for license
- Potential for collaboration with Navy researchers