Communications – electrical: acoustic wave systems and devices – Distance or direction finding – By combining or comparing signals
Reexamination Certificate
1999-12-02
2001-10-16
Lobo, Ian J. (Department: 3662)
Communications, electrical: acoustic wave systems and devices
Distance or direction finding
By combining or comparing signals
C367S131000, C342S378000, C455S067150
Reexamination Certificate
active
06304515
ABSTRACT:
BACKGROUND-FIELD OF THE INVENTION
In general, the present invention relates to the utilization of auto-correlation and cross-correlation functions in methods for detecting and communicating data from one or more sources of a primary signal when the primary signal has a waveform that is not known at one or more receivers at the outset.
BACKGROUND-DESCRIPTION OF PRIOR ART
The energy detector is the traditional method for detecting signals of uncertain emitted waveform (Whalen,
Detection of signals in noise
, Academic Press, New York, pp. 411, 1971). If the waveform was known, one could use the matched filter to enhance the detection of the signal (North, RCA Laboratory Report PTR-6C; (reprinted in Proc. I.E.E.E. 51, 1016-1027, 1963), 1943). However without knowledge of the waveform, the energy detector is the only prior alternative. The energy is the sum of the squares of samples from the received signal. If the energy changes by more than noise alone would likely allow, one decides that one or more signals are present, with some accompanying probability of error.
The main deficiency of the energy detector is that is probabilities of detection may not be as large as desired.
The prior-art most closely related to the present invention is discussed by Spiesberger in a U.S. Pat. No. 6,160,758. In that invention, the relative travel times of multipath are estimated using a plurality of input channels (claims
1
,
7
,
9
). Claim
10
in that application discusses a means for blindly estimating both the lags and amplitudes of multipath from a plurality of channels. In that invention, a set of new correlation equations are used to solve for the relative travel times and amplitudes of the multipath. The invention is U.S. Pat. No. 6,160,758 does not provide a method for estimating the relative travel times or amplitudes of multipath when data are taken on only one channel.
SUMMARY
The matched-lag filter provides methods for detecting multipath signals from one or more sources collected at one or more receivers with significantly higher probabilities of detection than can be obtained with the energy detector. As a result, the matched-lag filter offers a new way to detect multipath signals, and it offers a new way to communicate information in multipath conditions. The matched-lag filter also offers a new way for blindly estimating the amplitudes and relative travel times of multipath between a source of primary signal and at least one input channel.
OBJECTS AND ADVANTAGES
In order to detect signals with higher probabilities of detection than previously possible, a new method is developed for detecting broadband signals of unknown waveform that travel to one or more receivers along many paths. The theory has applications to acoustic and electromagnetic wireless communications, and to the detection of signals from calling animals, ships, vehicles, nuclear blasts, seismic events and other broadband phenomenon of acoustic and electromagnetic origin. Simulations of this theory in this paper indicate that it is able to detect signals with significantly larger probabilities of detection than the alternative method based on the energy of the signal at the receivers (Whalen, 1971).
The method is developed as a binary detection problem where the data consist of either noise or signal plus noise. It is assumed that the signal travels along an uncertain number of paths to each receiver because of reflections or refraction within the environment. The amplitude and emission time of the transmitted signal may be unknown, as may be the travel times of the paths. The lack of information about a library of transmitted waveform shapes may preclude the use of detection methods based on the matched filter (North, 1943). The method does not require a model for the propagation of waves in the media. If such information is provided, such as from matched-field processing (Bucker, Journal Acoust. Soc. Am. 59, 368-373 1976), it can be used to enhance the probability of detection. The signal's bandwidth must be sufficiently wide so that some paths arrive at intervals exceeding the inverse bandwidth of the signal. It is assumed that the variance of the noise is imperfectly known, either because one does not know when the signal is on or off, or because the noise is not stationary, or because the variance is estimated from the data.
Signals that cannot be reliably modeled may occur when acoustic or electromagnetic waves propagate above ground where the locations of boundaries such as the ground, trees, rocks, buildings etc. are unknown (Paulraj and Papadias,
IEEE Signal Processing Magazine
, 14, no. 5, 49-83, 1997; Spiesberger, Journal Acoust. Soc. Am., 106, 837-846, 1999), or when transient or long-lived broadband sounds propagate in water with complicated bathymetry, or when electromagnetic waves propagate through the ionosphere (Price, IRE Transactions on Information Theory, IT-2, no. 4, 125-135, 1956).
Signals that can be reliably modeled may occur when low frequency acoustic waves propagate through deep water, or through the Earth following earthquakes or nuclear blasts. The detection of all these signals has been of interest for a long time.
In the new method, use is made of the auto-correlation function when there is one receiver. For more than one receiver, all the auto- and cross-correlation functions are used. The use of all these correlation functions for detecting the signal appears to be novel. They offer a natural starting point because they provide the standard gain obtained with a matched filter (North, 1943) when the signals from the paths are partially coherent with one other.
Perhaps the most distinguishing feature of the new receiver is its use of the physically possible arrangements of signal lags in the auto- and cross-correlation functions. Multipath signals have been detected using arrival patterns that are modeled stochastically, such that the auto-correlation function of the multipath signals is assumed to be known ahead of time (Price, 1956; Davis, Transactions I.R.E., PGIT-3, 52-62, 1954; Youla, Transactions I.R.E., PGIT-3, 90-105, 1954; Middleton, I.R.E. Transactions on Information Theory, IT-3, 86-121, 1957). This assumption allows signals to occur at any lags in the auto-correlation function, but indeed there are many lag arrangements that are forbidden. Take, for example, three paths arriving at a receiver. There are at most 3(3−1)/2=3 signal-related peaks at positive lags in the autocorrelation function (Spiesberger, Journal Acoustical Society of America, 100,
t(2)
Occupied Lags
1
1
4
5
2
2
3
5
3
2
3
5
4
1
4
5
910-917, 1996). The lags of these peaks must satisfy the “lag-equations” (Spiesberger, 1996),
&tgr;
ii
(m,n)=t
i
(m)−t
i
(n);m>n≧1, (1)
where the travel time for path m at reciever i is t
i
(m). It is impossible to have lags 2, 4, and 5 simultaneously occupied by signals where sample times are counted by their sample number. To see this, the difference in sample number between the third and first paths must be 5, to yeild a signal at lag 5. This leaves one additional path which could arrive at sample numbers 1, 2, 3, or 4. None of theses possibilities yields signals at lags 2, 4, and 5 (Table I). There are also forbidden arrangements of signals in auto- and cross-correlation functions. Lag-equations have been found which describe the allowance arrangements of signals in these functions when there are two or more recievers (Spiesberger, Journal of the Acoustical Society of America, 104, 300-312, 1998).
So detecting the signals described in the first paragraph can involve more than looking at signal-to-noise ratios. Indeed, it is shown that the allowed lag arrangements in auto- and cross-correlation functions can be used to design a “matched-log filter” (Spiesbergher, Proc. 1999 I.E.E.E. International Conference on on Acoustics, Speech, and Signal Processing, III, 1189-1192, 1999) and thereby increase the probability of detection compared with conventional receivers which base their decision on the received energy.
Based upon the
Lobo Ian J.
Spiesberger John Louis
LandOfFree
Matched-lag filter for detection and communication does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Matched-lag filter for detection and communication, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Matched-lag filter for detection and communication will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2613450