Communications – electrical: acoustic wave systems and devices – Receiver circuitry
Reexamination Certificate
2001-01-12
2002-07-09
Lobo, Ian J. (Department: 3662)
Communications, electrical: acoustic wave systems and devices
Receiver circuitry
C367S901000, C367S124000
Reexamination Certificate
active
06418083
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention relates to a signal processing method for separating signals from noise and clutter signals, and more specifically to a signal processing method which identifies excess phase or amplitude shifts corresponding to noise and clutter signals.
Most signal processing technology used to differentiate between signals and noise and clutter concentrates on filtering higher amplitude inputs (assumed to be signals) from lower amplitude inputs (assumed to be noise). However, this method is effective only if the signal has a greater amplitude than the surrounding environmental noise. Signals with amplitudes approximately the same or less those that of the environment are extremely difficult to distinguish from noise.
Target detection and identification is very important in many applications, including the fields of underwater acoustics, radar processing, and infrared detection. As the signatures of mechanical systems decrease, signal processors which rely on identifying a signal of interest based on amplitude of the received signal alone lose their capability to distinguish between the signal of interest and noise. In this environment, increases in the signal to noise ratio of the signal processors become even more critical.
One method for increasing the signal to noise ratio of sinusoidal signals in random noise using a phase fluctuation approach is found in Wagstaff et al U.S. Pat. No. 6,353,578, which is incorporated herein in its entirety. Another method is described in U.S. Pat. No. 5,732,045, Fluctuations Based Digital Signal Processor Including Phase Variations issued to Wagstaffet al on Mar. 24, 1998, which is incorporated herein in its entirety. Another method is described in Wagstaffet al, U.S. Pat. No. 6,356,510, incorporated herein in its entirety. These phase based processors rely on an estimate of the excess phase rotation exhibited by a time series of data. The basis for these phase fluctuation processors is briefly described below.
The polar coordinate diagram of
FIG. 1
illustrates the phase relationship between three consecutive samples of a complex vector quantity R
i−2
, R
i−1
, and R
i
, which have amplitudes r
i−2
, r
i−1
, and r
i
and phase angles &thgr;
i−2
, &thgr;
1−1
, and &thgr;
i
. Phase alignment of consecutive samples of the complex vector quantity relies on the expectation that for uniformly sampled data, the phase rotation of an i−2 th vector will have a uniform progression as it rotates to the next two sample positions, i−1 and i. A vector is said to have uniform progression if it has a constant angular velocity and zero angular acceleration. If the rotation rate is constant (has zero angular acceleration), the angular separation between the first two phase angles, &thgr;
i−2
−&thgr;
i−1
, should be equal to the angular separation between the second two phase angles, &thgr;
i−1
−&thgr;
i
. The expected phase angle of the i th vector, &thgr;′
i
will be equal to the second phase angle, &thgr;
i−1
, plus the angular separation between the first and second phase angles, &thgr;
i−1
-&thgr;
i−2
(and &thgr;′
i
=&thgr;
i−1
+&thgr;
i−1
−&thgr;
i−2
=2&thgr;
i−1
−&thgr;
i−2
). If the rotation rate is not constant, the fluctuation in the phase angle &thgr;
i
of the vector R
i
will result in an excess phase rotation, &PHgr;
i
, which is equal to &thgr;
i
−&thgr;′
i
, the amount that the actual phase angle &thgr;
i
exceeds the expected phase angle &thgr;′
i
.
A phase fluctuation based processor uses estimates of the excess phase rotation, &PHgr;
i
, to measure the amount of phase fluctuation in a set of data. Note that the excess phase rotation is proportional to the angular acceleration. There are several methods to estimate the excess phase fluctuation, one of which is as follows:
If
R
i
=r
i
e
j&thgr;i
=r
i
Cos &thgr;
i
+jr
i
Sin &thgr;
i
=r
i
Z
i
, Equation (1)
where Z
i
is the i th unit phasor, then &PHgr;
i
, the excess phase rotation of the complex vector R
i
, can be expressed as
&PHgr;
i
=&thgr;
i
−&thgr;′
1
=&thgr;
i
−2&thgr;
i−1
+&thgr;
i−2
Equation (2)
If &PHgr;
i
is within the domain −&pgr; to +&pgr;, then &PHgr;
i
can be estimated by &PHgr;
i
=&thgr;
i
−&thgr;′
i
=&thgr;
i
−2&thgr;
i−1
+&thgr;
i−2
. The restriction to the range of −&pgr; to +&pgr; is necessary only when the values of &PHgr;
i
are used directly by themselves in later processing steps, for example when several values of &PHgr;
i
are averaged. When the value of &PHgr;
i
is transformed to a trigonometric or other function which is then used in later processing steps (such as when later processing steps compare values of cos &PHgr;
i
) then &PHgr;
i
is not restricted to a particular range.
There are other methods which may be used for estimating the excess phase rotation &PHgr;
i
. One is to assume a uniform progression of phase angles of one degree per sample i for the entire data set in a given frequency bin, then to calculate the average squared differences between the quadrature components of the modified and original data. For example, an averaged squared value 1/N &Sgr; (sin (&thgr;
i
+1/180)−sin &thgr;
i
)) for i=1 to N may be calculated for the modified phase angle (&thgr;
i
+1/180 degrees). This process is repeated for two degrees per sample (&thgr;
i
+2/180 degrees), three degrees per sample (&thgr;
i
+3/180 degrees), and so on, through 179 degrees per sample. The accepted angle per sample is the one that gives the least average squared difference. The corresponding values of &PHgr;
i
are then the differences between the accepted modified phase angles and the original phase angles. A similar process could be used in which modified phase angles could be obtained by a phase-locked loop with corresponding differences giving the values of &PHgr;
i
.
The set of excess phase rotation data, estimated by one of the above methods, can then be used as the basis for several types of phase fluctuation-based signal processors. Although other phase fluctuation based signal processors (i.e. those described in U.S. Pat. Nos. 6,353,578, 6,356,510, 5,732,045) exist, a phase based processor which is even simpler and more sensitive to low amplitude signals than those currently in use would be extremely useful in signal processing and target detection applications.
There is also a need to improve the performance of currently used signal processors, whether they rely on discriminating between signal and noise based on a comparison between the acoustic power level of a signal and acoustic level of noise, or on another method. These processors could be greatly enhanced by the addition of processing steps which differentiate between signal and noise based on the magnitude of phase fluctuation in the received data.
A very sensitive phase fluctuation based processor is developed herein which is computationally efficient and can be added easily to other existing signal processors to increase the signal to noise ratio for faint signals.
SUMMARY OF THE INVENTION
An object of the invention is to provide a signal processor which filters high phase fluctuation signals from small phase fluctuation signals in a computationally efficient way.
An object of the invention is to provide a signal processor which filters large amplitude fluctuation signals from small amplitude fluctuation signals.
An object of the invention is to improve the signal to noise ratio of existing signal processors.
An object of the invention is to enhance a signal processor which is based on the signal amplitude, fluctuation of the signal amplitude, or other signal characteristic, by adding steps for measuring the phase fluctuation of a set of data.
An object of the invention is to improve the performa
Mobbs Jackson A.
Wagstaff Ronald A.
Becker Dorothy I.
Karasek John J.
Lobo Ian J.
The United States of America as represented by the Secretary of
LandOfFree
Phase fluctuation based signal processor does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Phase fluctuation based signal processor, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Phase fluctuation based signal processor will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2836979