Back to results
US 9,071,494 B2Granted
Systems and methods for fast and precise frequency estimation
AssigneeThe Aerospace Corporation
InventorRajendra Kumar
Life of this patent
Estimating the term…
Abstract
Systems and methods are provided for fast and precise estimation of frequency with relatively minimal sampling and relatively high tolerance to noise.
Claims29
A method, comprising:
- storing time series samples of an input signal, comprising a plurality of constituent signals, on at least one memory;
- iteratively shifting, by one or more processors, a center frequency of the received signal by a frequency offset to generate a frequency shifted input signal and decomposing the frequency shifted input signal into a first plurality of component signals;
- iteratively estimating, by the one or more processors, the average power of each of the first plurality of component signals;
- iteratively selecting, a second plurality of component signals comprising a predetermined number of the first plurality of component signals with the highest average power;
- iteratively determining, by the one or more processors, a first estimate of frequencies of each of the second plurality of component signals the input signal, the first estimate comprising a first predetermined number of frequencies;
- iteratively determining, by the one or more processors, a second estimate of the frequencies of each of the second plurality of component signals;
- iteratively estimating, by the one or more processors, an average power of each of the second plurality of component signals based at least in part on the second estimate of the frequencies of each of the second plurality of component signals;
- and
- selecting as final estimates of frequencies of the plurality of constituent signals, by the one or more processors and from the second plurality of component signals, the second plurality of component signals with the highest average power.
6 more elements
A system, comprising:
- a memory configured to store time series samples of an input signal, comprising a plurality of constituent signals;
- a frequency shifter to iteratively shift a center frequency of the input signal;
- a filter to decompose the shifted input signal into a first plurality of component signals;
- a first selector to select a second plurality of component signals from the first plurality of component signals by identifying those first plurality of component signals with the highest average power;
- a first stage estimator, implemented by one or more processors, and configured to determine a first set of frequency estimates associated with each of the second plurality of component signals;
- a second stage estimator, implemented by the one or more processors and comprising a plurality of adaptive Kalman filters, the second stage estimator configured to determine a second set of frequency estimates associated with each of the second plurality of component signals;
- a second selector to select as final estimates of frequencies of the plurality of constituent signals from the second plurality of component signals, the second plurality of component signals with the highest average power.
4 more elements
At least one non-transitory computer-readable medium comprising computer-executable instructions that, when executed by one or more processors, executes a method comprising:
- determining a first estimate of one or more frequencies of a first plurality of component signals of an input signal, the first estimate comprising a first predetermined number of frequencies corresponding to respective ones of the first plurality of component signals;
- estimating the average power of each of the first plurality of component signals;
- selecting a second plurality of component signals comprising a predetermined number of the first plurality of component signals with the highest average power;
- determining a second estimate of one or more frequencies of the second plurality of component signals, the second estimate comprising a second predetermined number of frequencies, wherein the second predetermined number of frequencies is less than the first predetermined number of frequencies;
- determining an error estimate associated with each of the second estimates of the second plurality of component signals of the input signal;
- determining, by the one or more processors, a third estimate of one or more frequencies of the second plurality of component signals of the input signal based at least in part on the determined error estimates.
3 more elements
Dependencies are inferred from the claim text.