System and method for blindly acquiring frequency hopped spread spectrum signals

Patent metadata

Field	Value
Publication number	US-9954714-B2
Application number	US-201615143289-A
Country	US
Kind code	B2
Filing date	Apr 29, 2016
Priority date	Apr 29, 2016
Publication date	Apr 24, 2018
Grant date	Apr 24, 2018

Abstract

A system and method for acquiring a frequency hopped spread spectrum (FHSS) signal with no prior knowledge about the FHSS signal. In example implementations, an RF signal is received at a receiver. The RF signal is converted into a stream of digital signal levels. Energy detections are identified in the stream of digital signal as possible hops of a FHSS signal. A feature set is blindly acquired for defining an FHSS signal from the energy detections. At least one waveform classification is generated based on the feature set. Energy detections are re-acquired from the RF signal based on the waveform classification.

First claim

What is claimed is:

1. A method for acquiring a frequency hopped signal with no prior knowledge about the frequency hopped signal, the method comprising:
receiving a radio frequency (RF) signal at a receiver;
converting the RF signal into a stream of digital signal levels;
identifying energy detections in the stream of digital signal levels as possible hops of the frequency hopped signal;
blindly acquiring a feature set for defining the frequency hopped signal from the energy detections by:
acquiring shape features from the energy detections by determining signal bandwidth, dwell, and absolute power and performing a shape maximum likelihood classification to determine a likelihood that the energy detections belong to a same feature set without comparing to any predetermined parameters;
acquiring frequency features from the energy detections by determining a signal center frequency, a frequency, a total bandwidth, and channel spacing for each energy detection and performing a frequency maximum likelihood classification to determine a likelihood that the energy detections belong to a same feature set without comparing to any predetermined parameters; and
identifying at least one feature set from the shape maximum likelihood classification and from the frequency maximum likelihood classification;

generating at least one waveform classification based on the at least one feature set;
directly acquiring energy detections from the RF signal based on the waveform classification; and
reporting the at least one waveform classification as the frequency hopped signal.

2. The method of claim 1 , wherein the step of blindly acquiring the feature set includes
measuring selected signal characteristics of the energy detections,
identifying features from the measured signal characteristics, and
determining a likelihood that identified features are correlated to form the feature set.

3. The method of claim 2 , wherein the features identified from measured signal characteristics are features selected from a group consisting of
detect power,
on-time period/phase,
off-time period/phase,
detect frequency and bandwidth,
direction of arrival, and
waveform frequency use and bandwidth.

4. The method of claim 1 , wherein the step of blindly acquiring the feature set comprises using a shape acquisition detector to measure signal characteristics corresponding to waveform shape characteristics.

5. The method of claim 1 , wherein the step of blindly acquiring comprises detecting using a frequency acquisition detector to measure signal characteristics corresponding to waveform frequency characteristics.

6. The method of claim 1 , wherein the step of blindly acquiring the feature set comprises using a time acquisition detector to measure signal characteristics corresponding to waveform time characteristics.

7. The method of claim 1 , further including
receiving RF signals at an N-antenna array, and
pre-filtering the RF signals received at the N antennas based on direction of arrival,
wherein the step of blindly acquiring comprises using a direction acquisition detector to detect a waveform direction of arrival.

8. The method of claim 1 , wherein the step of blindly acquiring the features set further includes
determining a probability density function (“PDF”) for each feature, and
utilizing the PDF of the feature to determine a likelihood that the feature set and other features are correlated.

9. The method of claim 8 , further including utilizing a plurality of weak classifiers to determine a likelihood that acquired features and the feature set are correlated.

10. The method of claim 9 , wherein the plurality of weak classifiers utilize Kalman filtering to compute a likelihood that an acquired feature belongs to the feature set.

11. The method of claim 1 , wherein the step of generating the at least one waveform classification includes
subsequently blindly acquiring features from the energy detections in the stream of digital signals, and
correlating acquired features and the feature set utilizing the shape maximum likelihood classifier and the frequency maximum likelihood classifier.

12. A system for acquiring a frequency hopped signal with no prior knowledge about the frequency hopped signal, the system comprising:
a receiver including at least one antenna for receiving an RF signal;
a digitizer, wherein the digitizer is configured to perform analog to digital conversion of the RF signal and as a result, convert the RF signal to a digital signal stream;
an energy detector, wherein the energy detector is configured to identify energy detections in the digital signal stream;
a processor; and
a non-transitory computer-readable medium storing processor-executable instructions that when executed by the processor are configured to perform functions for
a blind acquisition component configured to blindly acquire a feature set to define a frequency hopped signal from the energy detections by:
acquiring shape features from the energy detections by determining signal bandwidth, dwell, and absolute power and performing a shape maximum likelihood classification to determine a likelihood that the energy detections belong to a same feature set without comparing to any predetermined parameters,
acquiring frequency features from the energy detections by determining a signal center frequency, a frequency, a total bandwidth, and channel spacing for each energy detection and performing a frequency maximum likelihood classification to determine a likelihood that the energy detections belong to a same feature set without comparing to any predetermined parameters, and
identifying at least one feature set from the shape maximum likelihood classification and from the frequency maximum likelihood classification,

a full ensemble waveform classifier that is configured to generate at least one waveform classification based on the feature set,
a directed re-acquisition component configured to acquire energy detections based on the waveform classification, and
a report generator that is configured to report the at least one waveform classification as a frequency hopped signal.

13. The system of claim 12 , wherein the non-transitory computer-readable medium stores executable instructions for the blind acquisition component that, when executed by the processor, are configured to
measure selected signal characteristics of the energy detections,
identify features from the measured signal characteristics, and
determine a likelihood that identified features are correlated to form the feature set.

14. The system of claim 12 , wherein the features identified from measured signal characteristics are features selected from a group consisting of
detect power,
on-time period/phase,
off-time period/phase,
detect frequency and bandwidth,
direction of arrival, and
waveform frequency use and bandwidth.

15. The system of claim 12 , wherein the non-transitory computer-readable medium stores executable instructions for the blind acquisition component that, when executed by the processor, are operative to perform functions for a shape acquisition detector to measure signal characteristics corresponding to waveform shape characteristics.

16. The system of claim 12 , wherein the non-transitory computer-readable medium stores executable instructions for the blind acquisition component that, when executed by the processor, are operative to perform functions for a time acquisition detector to measure signal characteristics corresponding to waveform time characteristics.

17. The system of

Assignees

• Boeing Co

Inventors

• Soto Ely
• Rector James Chris

Classifications

• H04L1/245

  by using the properties of transmission codes · CPC title

• H04L43/028

  by filtering · CPC title

• H04L27/2666 Primary

  Acquisition of further OFDM parameters, e.g. bandwidth, subcarrier spacing, or guard interval length · CPC title

• H04B1/713

  using frequency hopping · CPC title

• H04B2001/71563

  Acquisition · CPC title