Communications – electrical: acoustic wave systems and devices – Signal transducers – Receivers
Reexamination Certificate
2003-06-25
2004-11-23
Lobo, Ian J. (Department: 3662)
Communications, electrical: acoustic wave systems and devices
Signal transducers
Receivers
C073S659000, C367S176000
Reexamination Certificate
active
06822929
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to acoustic spectrum analysis and, in particular, to a micro acoustic spectrum analyzer that uses one or more microfabricated resonator filters to provide frequency analysis of a sound signal.
BACKGROUND OF THE INVENTION
Acoustic spectrum analysis involves decomposing an input spectrum into different frequency components as a function of time. To be useful, the relative sound intensities (i.e., the rate of sound energy transmission per unit area) in the different frequency ranges must be determined also. Therefore, an acoustic spectrum analyzer provides the level of a sound signal as a function of frequency.
The analysis of sound in various frequency bands can be a valuable means to monitor industrial equipment and processes, detect security intrusions, or evaluate military threats. Such analysis can be provided by a sensor comprising an acoustic spectrum analyzer. In an industrial setting, the sensor could be used to diagnose machine health by comparing the sound emitted by a defective machine to the normal operating spectrum, or by picking up a characteristic sound that might be emitted by a worn bearing or belt. The sensor could also be used for access control, by recognizing a spoken word or phrase; or an alarm system, by recognizing characteristic sounds, like broken glass or a turning lock mechanism. In a military setting, the sensor could be used to recognize characteristic sounds of particular military vehicles in the air, on the land, or at sea.
In addition, there is a growing recognition that a large array of simple acoustic sensors may have advantages over a small array of more sophisticated sensors. For example, often the signal that clearly identifies a military or security threat is rapidly attenuated with distance. This is particularly true for pressure radiation sources at frequencies above the audio range. A simple sensor in close proximity to the threat could provide the best identification of the source. An array of such strategically located, miniaturized sensors could be used to monitor a large building or battlefield.
The standard approach to acoustic spectrum analysis is to use digital signal processing (DSP) to analyze the spectrum picked up and recorded by a microphone and recording system. DSP typically uses fast Fourier transform (FFT) methodology to quickly translate time-domain signals and provide a frequency-domain spectrum. However, DSP with an FFT analyzer requires extensive signal processing and considerable computing power to analyze the massive amounts of data acquired, which implies large physical size and power requirements. Because acoustic signals are by their nature nonstationary, this is particularly the case with real-time analysis. For example, current speech recognition systems require the computing power and data processing capabilities of a state of the art personal computer. Furthermore, if the computing power is remote from the sound source, the data communication overhead can be enormous. Even with the advances in modern radio frequency communications and computing technologies, it is not possible to do acoustic spectrum analysis in a truly miniature, low-power system. Therefore, in certain applications the power and size overhead of DSP make it impractical.
Alternatively, a simple acoustic spectrum analyzer can comprise one or more filters combined with a sound level meter. A real-time acoustic spectrum analyzer based on this approach relies on a bank of bandpass filters that pass selective frequencies, enabling the signal to be evaluated over a set of frequency bands simultaneously. Filters thereby enable the real-time frequency analysis of nonstationary signals and the more rapid analysis of stationary signals. FFT techniques can be used in combination with the filter set to extend the capabilities of the simple filter bank analyzer.
However, there remains a need for a low-power, low-cost, small-sized acoustic spectrum analyzer that can be used in applications were size, accessibility, and power availability are limited. The present invention provides a microfabricated acoustic spectrum analyzer that can fulfill this need in a low-power, miniature system. The micro acoustic spectrum analyzer combines novel microfabricated resonator technologies with state-of-the art signal processing and microphone technology to provide spectral analysis on a chip. The micro acoustic spectrum analyzer reduces the amount of data to be processed and exfiltrated through the use of selective resonant filtering of the input signal. Therefore, by doing the data analysis on-chip, much of the computational and data communication load of DSP is eliminated. The micro acoustic spectrum analyzer can be designed specifically for portability, size, cost, accuracy, speed, power requirements, use in a harsh environment, or whatever the application calls for.
In its simplest form, the micro acoustic spectrum analyzer of the present invention comprises a limited set of microfabricated resonators to provide signal analysis tailored to a specific application so that it “listens” selectively in the frequency bands of particular interest. This simple micro acoustic spectrum analyzer does not provide a full-blown, wide-band, high-resolution acoustic spectrum analysis system capable of recognizing any arbitrary sound. However, by recording and analyzing only the data of interest, the simple micro acoustic spectrum analyzer can dramatically reduce the data communication and processing overhead required to recognize a target sound, voice, or word. This targeted capability enables a significant reduction in size and power requirements that could be used in situations that preclude a full-blown analysis system. For example, the small size, low power, low cost of the on-chip acoustic spectrum analysis technology enables deployment of one or more sensors near the sound source. The sensor could then act as an intelligent input, or prefilter, to a remote, more capable digital signal processing system. With a sufficient number of resonator filters, the micro acoustic spectrum analyzer can provide analysis of a sound having a broader frequency range. This more sophisticated micro acoustic spectrum analyzer can be used for speech analysis and recognition, for example.
SUMMARY OF THE INVENTION
The present invention is directed to a micro acoustic spectrum analyzer for determining the frequency components of a fluctuating sound signal comprising a microphone to pick up the fluctuating sound signal and produce an alternating current electrical signal; at least one microfabricated resonator, each resonator having a different resonant frequency, that vibrate in response to the alternating current electrical signal; and at least one detector to detect the vibration of the at least one microfabricated resonator. The micro acoustic spectrum analyzer can further comprise a mixer to mix a reference signal with the alternating current electrical signal from the microphone to shift the frequency spectrum to a frequency range that is a better matched to the resonant frequencies of the at least one microfabricated resonator. The invention can further comprise a means for scanning and storing the detected vibrations and a pattern recognition processor to compare the detected vibrations to a library of profiles.
REFERENCES:
patent: 5640133 (1997-06-01), MacDonald et al.
patent: 5729075 (1998-03-01), Strain
patent: 5836203 (1998-11-01), Martin et al.
patent: 6012334 (2000-01-01), Ando et al.
patent: 6227054 (2001-05-01), Ando et al.
patent: 6236281 (2001-05-01), Nguyen et al.
patent: 6404304 (2002-06-01), Kwon et al.
patent: 6455980 (2002-09-01), Bernstein
Adkins et al., “Microfabricated Teeter-Totter Resonator”, patent application No. 10/436,597, filed May 12, 2003.
Stafford et al., “Multi-Tunable Microelectromechanical System Resonator”, patent application No. 10/443,951, filed May, 21, 2003.
Martin et al., “Flexural plate wave resonator excited with Lorentz forces,”J. Appl. Phys. 83(9), 4589 (1998).
M. Elwenspoek and R. Wiegerink,Mecha
Adkins Douglas R.
Anderson Larry F.
Butler Michael A.
Schubert W. Kent
Bieg Kevin W.
Lobo Ian J.
Sandia Corporation
LandOfFree
Micro acoustic spectrum analyzer does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Micro acoustic spectrum analyzer, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Micro acoustic spectrum analyzer will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3322224