Patent Grant
10280787
Any and all applications for which a foreign or domestic priority claim is identified in the Application Data Sheet as filed with the present application are hereby incorporated by reference under 37 CFR 1.57.
Field
This disclosure relates generally to systems and methods for monitoring rotating machinery, such as turbomachinery, using signals that have propagated from a transmitter to a receiver through a channel as waves in order to obtain information about the transmitter, the receiver, and/or the channel (including a target, such as turbomachinery equipment, located in the channel). More particularly, this disclosure relates to systems and methods for monitoring rotating machinery by performing coherent signal synthesis (at the transmitter) and/or analysis (at the receiver) to obtain information about the transmitter, receiver, and/or a frequency-selective channel, such as a multipath channel.
Description of the Related Art
The term turbomachinery generally describes a class of machines which are powered by, or harness energy from, a fluid (including liquids and gases). Turbomachinery can include, for example, turbines, which convert energy from a flowing fluid into rotary mechanical motion for performing work. Turbomachinery can also include compressors and fans, which use rotary mechanical motion to perform work on a fluid.
One important and ubiquitous type of turbomachinery is the gas turbine engine. Gas turbine engines are often used to provide thrust for airplanes or to power other types of vehicles or equipment. Generally speaking, a gas turbine engine includes a compressor with one or more stages which pressurize air. The pressurized air is then combined with fuel and combusted. The combustion generates a high temperature, high pressure flow of exhaust. A turbine is provided downstream and is used to harness energy from this exhaust flow. The turbine can in turn be used to power the compressor and other equipment, such as the fan in a turbofan engine.
Modern turbomachinery is often designed to satisfy a number of difficult operating requirements, including close mechanical tolerances, high temperatures, high pressures, high mechanical stresses, harsh operating environments, etc. Because of these difficult operating requirements, it would be desirable to have improved systems and methods for monitoring turbomachinery equipment and/or other types of machinery. Such monitoring can include testing, analyzing, characterizing, conducting failure detection and prediction, etc.
In some embodiments, a method for monitoring rotating machinery comprises: providing at least one transmitter antenna with access to at least a portion of the rotating machinery; providing at least one receiver antenna with access to the portion of the rotating machinery; obtaining at least one receiver signal resulting from at least one transmitter signal that has propagated from the transmitter antenna to the receiver antenna through the portion of the rotating machinery; forming at least a first signal pair which comprises a first receiver signal and a first transmitter signal, or first and second receiver signals which are obtained from spatially-separated receiver antennas, or first and second receiver signals which are attributable to different transmitter signals; determining amplitude and phase information of a plurality of frequency components for each signal in the first signal pair; determining a set of comparison values for the first signal pair by comparing respective frequency component phases and respective frequency component amplitudes of the signals in the first signal pair; and analyzing a characteristic of the rotating machinery using the set of comparison values. In some embodiments, the method further comprises coherently receiving first and second receiver signals and/or coherently synthesizing first and second transmitter signals. The method can also comprise controlling an operating condition of the rotating machinery based on the characteristic.
The rotating machinery comprises may be a gas turbine engine. The method can comprise positioning the transmitter antenna and the receiver antenna on opposite sides of a turbine stage of the rotating machinery, or on opposite sides of a compressor stage of the rotating machinery, or on opposite sides of a bypass fan of the rotating machinery, or with access to a bearing of the rotating machinery, or with access to a combustor of the rotating machinery, or with access to an exit nozzle of the rotating machinery.
In some embodiments, a system for monitoring rotating machinery comprises: at least one transmitter antenna configured to access to at least a portion of the rotating machinery; at least one receiver antenna configured to access to the portion of the rotating machinery; and a processor configured to obtain at least one receiver signal resulting from at least one transmitter signal that has propagated from the transmitter antenna to the receiver antenna through the portion of the rotating machinery; form at least a first signal pair which comprises a first receiver signal and a first transmitter signal, or first and second receiver signals which are obtained from spatially-separated receiver antennas, or first and second receiver signals which are attributable to different transmitter signals; determine amplitude and phase information of a plurality of frequency components for each signal in the first signal pair; determine a set of comparison values for the first signal pair by comparing respective frequency component phases and respective frequency component amplitudes of the signals in the first signal pair; and analyze a characteristic of the rotating machinery using the set of comparison values. The system can include receiver circuitry to coherently receive the first and second receiver signals, as well as transmitter circuitry to coherently synthesize first and second transmitter signals.
The transmitter antenna and the receiver antenna can be configured to be inserted into the rotating machinery from outside the machinery, or to be to be internally integrated with the rotating machinery. The rotating machinery may be a gas turbine engine.
The systems and methods described herein are useful for analyzing signals that have propagated from a transmitter to a receiver through a frequency-selective channel, such as a multipath channel, in order to determine information about the transmitter, the receiver, and/or the channel (including one or more targets located in the channel). As discussed further herein, the channel can include at least a portion of a piece of rotating machinery, such as turbomachinery. These systems and methods can take advantage of, for example, multipath propagation effects that cause modified versions of a transmitted signal to arrive at the receiver after having traversed the multipath channel. (Such multipath propagation effects are discussed with respect to
In the illustrated example, RF waves from the transmitter antenna T1 arrive at the receiver antenna R1 along a line of sight (LOS) pathway and two other multipaths M1 and M2 which result from the presence of the targets 130, 132. In some cases, the multipath effects introduced by the targets 130, 132 can be time-varying. For example, a target in the multipath channel can be physically moving or it can have some other time-varying characteristic which affects the RF waves received at the receiver. The collective response consisting of effects from the transmitter, the channel, and the receiver can be referred to as the system response, the system impulse response, the system transfer function, the time varying system impulse response, the time-varying system transfer function, etc.
In many applications, multipath signals are undesirable and are often considered to be an impairment. However, the systems and methods described herein can take advantage of multipath propagation effects (or other effects which occur in other types of frequency-selective channels) to detect changes in the propagation channel, including changes in one or more characteristics of the targets 130, 132. Multipath propagation effects can modify a transmitted signal in many ways, including by introducing (through scattering, reflection, refraction, diffraction, etc.) constructive or destructive interference, phase shifting, time delay, frequency shifting, and/or polarization changes to each multipath component. The systems and methods described herein can use techniques for identifying, measuring, and/or otherwise analyzing any of these effects, or others, to gain information about the multipath channel, including the targets 130, 132 located in the channel. It should be understood, however, that while various embodiments in this application are described in the context of multipath propagation channels, the systems and techniques described herein are also applicable to other types of frequency-selective channels. For example, the channel could be one in which one (or perhaps more) path(s) are themselves frequency-selective, such as a frequency-selective medium or a frequency selective surface reflection.
In addition, besides being used to gain information about the channel (including one or more targets located in the channel), the systems and methods described herein can also be used to gain information about the transmitter and/or the receiver. For example, the systems and methods discussed herein can be used to identify or characterize changes in the polarization state of the transmitted signals, changes in the orientation or location of transmitter antennas, changes in a combination of signals from multiple transmitter antennas (e.g., changes in the amplitude and/or phase weighting factors applied to multiple transmitted signals), changes in the relative delays between transmitted signals, etc. Similarly, the systems and methods discussed herein can be used to identify or characterize similar effects at the receiver. Any of these effects impacting the system response can be identified, measured, and/or otherwise analyzed to gain information about the transmitter, the receiver, and/or the channel (including the targets 130, 132 located in the channel).
Thus, the systems and methods described herein can characterize not only the channel but also the transmitter and/or receiver. For example, if the transmitter and receiver are fixed, then the measured signals can be used to characterize changes in the channel. But for a fixed channel and a fixed receiver, the measured signals can characterize changes in the location and/or properties of the transmitter. Similarly, for a fixed transmitter and channel, the received signals can characterize changes in the location and/or properties of the receiver. Or, in general, the measured signals can contain information about transmitter effects, channel effects, and receiver effects (which effects may or may not be separable).
The received signal(s) represent the convolution of the transmitted signal(s) with the channel, and hence is/are a function of the transmitted signal. When the transmitted signal(s) is/are known, that knowledge can be used by the receiver to estimate the system response, typically with greater accuracy than if the transmitter signal is not known. This capability has an advantage of limiting the impacts due to the specific waveforms that are transmitted, especially those exhibiting any time-varying spectral properties.
The system 200 illustrated in
The transmitter 210 transmits a signal ST1x of bandwidth BW centered at RF frequency f0. One way to accomplish this is to generate a baseband signal of bandwidth BW and to up-convert this signal to an RF carrier frequency f0. The resulting signal may be transmitted through the transmitter antenna T1. Alternatively, the transmitter can transmit a signal consisting of at least two tones that are spaced apart in frequency, or the transmitter can sweep the frequency of a tone or pulse an RF tone. In some embodiments, a signal having a bandwidth BW centered at the RF frequency f0 can be directly generated using digital signal processing followed by digital-to-analog conversion. Other methods of signal generation are also possible.
The transmitted signal emitted from the transmitter antenna T1 begins propagating through the multipath channel as x-polarized RF waves across the full range of frequencies comprising the bandwidth BW of the transmitted signal. In the case considered, the multipath channel includes one or more targets 230 which introduce multipath contributions at the receiver 220, which can result in a frequency-selective vector propagation channel (i.e., a frequency-selective channel for at least one of the polarization modes) if path delays among the components exhibit sufficient spread. The receiving antenna R1 detects orthogonally-polarized channel-modified versions of the transmitted RF signal. The signal SR1u represents the u-polarized component of the detected signal, whereas the signal SR1v represents the v-polarized component. These orthogonally-polarized signals can be processed at the receiver 220 in order to determine information about the transmitter, the channel, and/or the receiver. If the transmitter and receiver are fixed, for example, then the received signals can be used to detect and characterize changes in the multipath channel. This is discussed in U.S. Patent Publication 2013/0332115, the entire contents of which are hereby incorporated by reference in this disclosure.
In some embodiments, the receiver 220 down-converts the received RF signals and performs analog-to-digital conversion. The down-converted signals can be represented in any suitable form, including as in-phase and quadrature signal components. The down-converted SR1u and SR1v signals can be analyzed sub-band by sub-band. For example, the receiver 220 can perform an N-point fast Fourier transform (FFT), or other suitable transform, to convert the signals into N bins in the frequency domain. Each of these frequency bins can be considered as a sub-band (also referred to as a sub-frequency or sub-carrier). If, for example, the originally-transmitted baseband signal has a bandwidth of 20 MHz, the received SR1u and SR1v signals can divide the 20 MHz bandwidth into any number of sub-bands which can then be considered independently, or in combination, to analyze the transmitter-channel-receiver system as a function of frequency.
In some embodiments, the receiver 220 calculates the polarization for each sub-band by using the frequency-domain representations of the baseband SR1u and SR1v signals to calculate a Jones vector or Stokes parameters (which can be obtained by calculating the Jones coherency matrix). These calculations are known in the art and examples are provided in U.S. Patent Publication 2013/0332115, which are incorporated herein by reference. When calculated using signals from a dual polarization (orthogonally-polarized) antenna, the result of these computations is polarization state information. The polarization information may be computed for each sub-band of the down-converted baseband signals received at the antenna R1. The polarization can be measured in a relative sense, or, if the orientation of the receiver antenna R1 is known, in an absolute sense. Polarization statistics, such as the degree of polarization can also be measured for the entire signal. Alternatively, repeated measurements of the state of polarization for each sub-band can be used to characterize the degree of polarization associated with the sub-band.
The polarization state information characterizes the polarization mode dispersion—the frequency-dependency of the polarization mode shifting—caused by the channel or other factors. The polarization values (e.g., the Stokes parameters) for each sub-band can be normalized, where the S1, S2, and S3 Stokes parameters are scaled to form a vector of unit magnitude, depending upon whether or not the signal has a unity degree of polarization. (Using a small enough sub-band spacing will generally yield a degree of polarization near unity in each sub-band.) The resulting polarization values may be plotted on or about a Poincaré sphere as a visualization aid. For example, the normalized S1, S2, and S3 Stokes parameters for each sub-band can be taken as coordinates and plotted on the Poincaré sphere (which has a unit radius) as a point. Each location on the Poincaré sphere corresponds to a different polarization state. When the Stokes parameters for multiple sub-bands are plotted, the result is a locus of points which can be referred to as a polarization mode dispersion (PMD) curve. As discussed in U.S. Patent Publication 2013/0332115, PMD curves can be analyzed to determine information about the multipath channel. They may also provide information about any other type of frequency selective channel or about any portion of the transmitter-channel-receiver system.
While normalization of the S1, S2, and S3 Stokes parameters to a unit vector may be advantageous in some embodiments, in other embodiments retaining the amplitude information in the parameters is desirable, in which case the S0 value will be maintained along with S1, S2, S3. The unnormalized parameters S1, S2, and S3 taken from the full Stokes vector [S0 S1 S2 S3] can also be plotted in 3D space, but will not, in general, be confined to a locus that resides on a unit sphere, yet the resulting curve may still be analyzed to determine information about the transmitter-channel-receiver system. Also, it may also be useful to retain RF phase information of the signals used in the formation of the Stokes parameters.
While
The transmitter 310 transmits a signal ST1x with a bandwidth BW centered at an RF frequency f0 via the antenna T1. The transmitter signal can be generated in any way disclosed herein, for example. The signal propagates through a frequency-selective channel, such as a multipath channel, with one or more targets 330 that create a frequency-selective response at the receiving antennas R1, R2. The channel, for example, can cause different modified versions of the transmitted signal ST1x to be received at the spatially-separated receiving antennas R1, R2. The signal SR1 represents the signal received at R1, whereas the signal SR2 represents the signal received at R2. The receiver 320 can down-convert these signals and perform analog-to-digital conversion. As discussed further herein, the received signals SR1 and SR2 can be coherently received (e.g., coherently sampled and processed). In addition, the two receiver channels for these signals can be phase and/or gain matched.
Once, the SR1 and SR2 signals are down-converted and sampled, the frequency component phases and amplitudes of the baseband SR1 and SR2 signals can be compared. This can be done in the time domain (e.g., via a filter bank) or in the frequency domain. For example, each of the received signals can be converted into the frequency domain using an N-point FFT operation. This operation divides the bandwidth of each of the down-converted SR1 and SR2 signals into N frequency bins. The respective amplitudes and phases of the frequency components of the SR1 and SR2 signals can then be compared for each sub-band. For example, the amplitudes of the frequency components of one of the signals can be compared to those of the other by calculating differences between the respective amplitudes or ratios of the amplitudes. Similarly, the phases of the frequency components of one of the signals can be compared to those of the other by calculating differences between the respective phases. These are just some examples of computations which can be performed to compare the respective amplitudes and/or phases. Many others are also possible. For example, in some embodiments, the respective amplitudes and phases of the frequency components of the SR1 and SR2 signals can be compared by calculating a Jones vector or Stokes parameters (normalized or unnormalized) for each sub-band using the SR1/SR2 signal pair. Other mathematical computations can also be used to compare the phases and/or amplitudes of the frequency components of the two signals.
If the SR1 and SR2 signals had been obtained from a dual polarized antenna, then the results of this computation would be polarization information (as already discussed above with respect to
As just mentioned, the respective frequency component amplitudes and/or phases of each of these signal pairs can be compared (e.g., for each sub-band). (As already disclosed, one example of the comparison values that can be calculated are the Stokes parameters for each sub-band of each signal pair. Stokes parameters (S0, S1, S2, and S3) for each sub-band can be calculated according to the following equations: S0=(Y1·Y*1)+(Y2·Y*2); S1=(Y1·Y*1)−(Y2·Y*2); S2=(Y1·Y*2)+(Y2·Y*1); and S3=j(Y1·Y*2)−j(Y2·Y*1), where Y1 is a complex number with amplitude and/or phase information for a first signal in the pair of signals being compared and Y2 is a complex number with amplitude and/or phase information for a second signal in the pair of signals being compared.) The phases can be measured only in a relative sense with respect to one another or with respect to a local oscillator at the receiver 320. Alternatively, and/or additionally, the phases can be measured with respect to a phase reference (e.g., a local oscillator) at the transmitter 310. Frequency dispersion statistics (likened to degree of polarization) can be determined for each sub-band. Other computations for estimating the same or similar information can be calculated from power measurements as described in Pratt et al., “A Modified XPC Characterization for Polarimetric Channels,” IEEE Transactions on Vehicular Technology, Vol. 60, No. 7, September 2011, p. 20904-2013. This reference describes polarization characterizations, but the same techniques can be applied to the signals pairs disclosed herein even though they will not result in polarization information. This reference is therefore incorporated by reference herein in its entirety for its disclosure of such analysis techniques.
In some embodiments, the receiver 320 can include more than two receiving antennas to obtain additional receiver signals. In addition, in some embodiments, the system 300 architecture can be reversed from what is shown and can instead include two or more transmitter antennas for sending two or more transmitter signals and only one receiver antenna for obtaining a receiver signal. (In embodiments with two or more transmitter signals, the transmitter signals can be coherently synthesized, as discussed further herein.) Or the system 300 could include two or more transmitter antennas (for sending two or more transmitter signals) and two or more receiver antennas (for obtaining two or more receiver signals). In any case, all of the resulting signal pairs can be used to analyze the system, as disclosed herein.
The transmitter 410 transmits a signal ST1x with a bandwidth BW centered at an RF carrier frequency f0 via the antenna T1. The signal ST1x can be generated using any technique disclosed herein or any other suitable technique. The channel can include one or more targets 430 which create one or more signal paths to the receiving antennas R1, R2. These signal paths result in frequency-selective propagation effects that typically cause different modified versions of the transmitted signal ST1x to be received at the spatially-separated dual polarized receiving antennas R1, R2. The first receiving antenna R1 detects orthogonally-polarized components of channel-modified versions of the transmitted RF signal. The signal SR1u represents the u-polarized component of the detected signal at the first receiving antenna R1, whereas the signal SR1v represents the v-polarized component. The second receiving antenna R2 likewise detects orthogonally-polarized components of channel-modified versions of the transmitted RF signal. The signal SR2u represents the u-polarized component of the detected signal at the second receiving antenna R2, whereas the signal SR2v represents the v-polarized component.
The orthogonally-polarized signal components from each of the receiving antennas R1, R2 can be processed at the receiver 420 in order to determine information about the transmitter-channel-receiver system. The receiver 420 can down-convert these signals and perform analog-to-digital conversion. As discussed further herein, the received signals SR1u, SR1v, SR2u, and SR2v can be coherently received (e.g., coherently sampled and processed). In addition, the four receiver channels for these signals can be phase and/or gain matched. Once, the SR1u, SR1v, SR2u, and SR2v signals are down-converted and sampled, the frequency component phases and amplitudes of various signal pairs can be compared. The different signal pairs are described below with respect to
Each of the received signals SR1u, SR1v, SR2u, and SR2v can be converted into the frequency domain using an N-point FFT operation. This operation divides the bandwidth of each of the baseband SR1u, SR1v, SR2u, and SR2v signals into N frequency bins. The respective frequency component amplitudes and phases of the various pairs of signals can then be compared for each sub-band using any calculation discussed herein or any other suitable calculation. In some embodiments, the respective frequency component amplitudes and phases for a particular signal pair can be compared by, for example, calculating a Jones vector or Stokes parameters (normalized or unnormalized) for each sub-band. Additionally absolute phase and amplitude information and statistics can also be measured.
Additional information about the transmitter-channel-receiver system can be obtained by also comparing respective frequency component phases and/or amplitudes from signals detected at different antennas. A total of four signal pairs can be formed to make these “cross-antenna” comparisons. These are signal pairs 3-6 in the table shown in
The first six signal pairs in the table shown in
As discussed herein, the respective frequency component phases and/or amplitudes for each of the signal pairs from the table shown in
In some embodiments, the receiver 420 can include more than two dual polarized receiving antennas to obtain additional receiver signals. In addition, in some embodiments, the system 400 architecture can be reversed from what is shown and can instead include two or more transmitter antennas (which can be spatially-separated and/or dual polarized) for sending two or more transmitter signals and only one receiver antenna (which can be dual polarized) for obtaining a receiver signal. Or the system 400 could include two or more transmitter antennas (for sending two or more transmitter signals) and two or more receiver antennas (for obtaining two or more receiver signals). In any case, all of the resulting signal pairs can be used to analyze the system, as disclosed herein.
The transmitter 510 includes two waveform generators 504a, 504b that can respectively provide baseband waveforms ST1x and ST1y that are coherently synthesized and centered at a carrier frequency f0 and transmitted via the transmitting antenna T1. The waveform generators 504a, 504b can provide any of the following waveforms: single tone continuous wave, wideband noise, band-limited noise, chirp, stepped frequency, multi-tone, pulses, pulsed chirps, orthogonal frequency division multiplexing (OFDM), binary phase shift keying (BPSK), linear FM on pulse (LFMOP), etc. It should be understood, however, that these are just example waveforms and that a wide variety of other waveforms can also be used, including any desired arbitrary waveform that may be suited to a given application. Each of the waveform generators 504a, 504b can operate independently and can provide different waveforms at any given time. In some embodiments, the transmitted signals can be scaled and/or phase-shifted versions of one another. For example, when using a dual-polarized transmit channel, controlling the relative phase and amplitude between the orthogonally-polarized channels leads to control over the transmitted polarization state. In other embodiments, it is also possible to generate time-delayed signals, each with a controlled relative scaling and/or shift between the orthogonally-polarized channels, for example to intentionally induce dispersion.
The baseband waveforms produced by the waveform generators 504a, 504b are provided to up-converters 502a, 502b to be centered at an RF carrier frequency f0. The RF carrier frequency is provided by the local oscillator 508. The carrier frequency is fed from the local oscillator 508 to the up-converters 502a, 502b via signal lines 506a, 506b. In some embodiments, the signal lines 506a, 506b are matched signal lines so as to maintain the phase coherency of the carrier frequency at the up-converters 502a, 502b. As shown in
As just discussed, in some embodiments, the transmitted signals are coherent. Phase information can be preserved between the various transmitter signals. One way to achieve coherency between the transmitted signals is to share a common local oscillator 508 used in the up-conversion processing. A common local oscillator can be advantageous in a multichannel transmitter because any impairments in the local oscillator may affect all channels relatively equally, thus not substantially affecting relative channel-to-channel comparisons. In some instances, control over the local oscillator phase may be advantageous, for example to assure that the starting phase reference for each transmitted signal is substantially identical (or if not identical then known so that the phase difference between transmitted signals can be compensated). In some embodiments, the transmitter can advantageously achieve precise control of the phase, amplitude, sampling, and frequency among the various generated signals used at the transmitter. Further, in some embodiments, the phase noise of the local oscillator 508 is negligible such that energy of a desired signal in one sub-band coupling to an adjacent sub-band is significantly less (e.g., two or more orders of magnitude less) than the signal being detected in that adjacent band.
In addition, in some embodiments, each signal channel in the transmitter can be substantially phase and gain matched with the others. In order to achieve this matching, compensation circuits can be included. For example, if the transmitter includes different amplifier circuits in each channel, then depending upon the transmit signal and the non-linear behavior of the amplifier in each channel, it may be possible for asymmetrical signal distortion to occur (e.g., the effects on one channel are not identical to the other channels). Such behavior could be detrimental to a coherent, matched system, and so compensation circuits can be used to reduce or minimize phase and gain mismatches in the channels.
Although the transmitter 510 in
In some embodiments, the transmitted signals ST1x and ST1y are advantageously separable. This means that the transmitted signals ST1x and ST1y have the property that they can be distinguished from one another by the receiver 520. For example, the different signals generated at the transmitter may be approximately orthogonal in some sense so that the signals can be separated at the receiver with little crosstalk among the signals. The multiple signals generated at the transmitter can be sent using a different signal on each antenna, or by using different linear combinations of multiple antennas to transmit each signal. In addition, the transmitted signals can employ, for example, a cyclic prefix to help reduce inter-symbol interference (non-orthogonal subcarriers).
The separability property of the transmitted signals can be achieved in several different ways, including, for example, through the use of time division multiplexing, frequency division multiplexing, and/or code division multiplexing. Methods based on eigendecomposition or singular value decomposition can also be used. Other methods may also be possible. In the case of time division multiplexing, the signals ST1x and ST1y can be transmitted during different time slots such that the receiver can distinguish the response of each of the receiving antennas to each of the transmitted signals. However, in many cases the system 500 is used to detect a time-varying property of a multipath channel. Therefore, it may be desirable to transmit both of the signals ST1x and ST1y at the same or overlapping times in order to more completely characterize the time-varying property. This is particularly true if the variations being monitored occur on a timescale that is short as compared to the length of the time slots for the transmitted signals. In cases where it is desirable that the signals ST1x and ST1y be transmitted at the same time (or at time periods which overlap), then frequency division multiplexing, code division multiplexing, eigendecomposition, singular value decomposition, and/or other methods can be used.
The transmitter 510 transmits the separable baseband signals ST1x and ST1y, up-converted to the RF carrier frequency, via the antenna T1. The ST1x signal is transmitted via the x-polarized component of the transmitting antenna T1, while the ST1y signal is transmitted via the y-polarized component of the transmitting antenna. (It is also possible that the signals can be transmitted using different weighted combinations of the x- and y-polarization modes.) The frequency-selective channel (in this example, a multipath channel) includes one or more targets 530 which create multiple signal paths to the receiving antennas R1, R2. These multiple signal paths result in multipath propagation effects that cause different modified versions of the separable transmitted signals ST1x and ST1y to be received at the spatially-separated, dual polarized receiving antennas R1, R2.
The first receiving antenna R1 detects orthogonally-polarized components of the received RF signals. The signal notation SR1uT1x can be used to represent the u-polarized component of the detected signal at the first receiving antenna R1 due to the transmitted signal ST1x, while the signal SR1vT1x represents the v-polarized component of the detected signal at the first receiving antenna R1 due to the transmitted signal ST1x. In this notation, for any given received signal the subscript indicates the receiving antenna and polarization channel whereas the superscript indicates the transmitted signal which excited that particular received signal. Using this notation, the u- and v-polarization components detected at R1 due to the transmitted signal ST1y can be written as SR1uT1y and SR1vT1y, respectively. Similarly, the u- and v-polarization components detected at R2 due to the transmitted signal ST1x can be written as SR2uT1x and SR2vT1x, respectively. And the u- and v-polarization components detected at R2 due to the transmitted signal ST1y can be written as SR2uT1y and SR2vT1y, respectively.
These signals can be processed at the receiver 520 in order to determine information about the transmitter-channel-receiver system. Part of the processing that can be performed by the receiver 520 is separating the signal responses at each of the four antenna inputs which are attributable to each of the transmitted signals ST1x and ST1y. For example, the response at the u-polarization component of the first receiver antenna R1 will, in general, consist of a superposition of channel-modified versions of the transmitted signals ST1x and ST1y transmitted at both the x- and y-polarizations, respectively. The same will generally be true of the response at the v-polarization component of the first receiving antenna R1 and of the u- and v-polarization components of the second receiving antenna R2. The receiver 520 can perform signal separation operations to isolate the response at each receiver input that is attributable to each of the transmitted signals.
In the case where the transmitted signals ST1x and ST1y are made separable using frequency division multiplexing (as shown in
In summary, for cases where the transmitter 510 transmits multiple signals, the detected response at each input port of the receiver 520 will in general consist of the superposition of transmitter-, receiver-, and/or channel-modified versions of each of the multiple transmitted signals (especially if the multiple transmitted signals are coincident in time). The signal separation operations performed by the receiver 520 isolate these superimposed signals in order to determine the individual response at each polarization component of each receiver antenna which is attributable to each transmitted signal. In the case of the system 500 in
The receiver 520 can down-convert the SR1uT1x, SR1vT1x, SR1uT1y, SR1vT1y, SR2uT1x, SR2vT1x, SR2uT1y, and SR2vT1y signals and perform analog-to-digital conversion. This is done using the down-converters 522a-d and the analog-to-digital converters 524a-d. Each of these components can be connected to, and controlled by, a common local oscillator 528 and/or clock signal (as applicable depending upon the circuitry) in order to maintain consistent phase and/or timing references. For example, the signals can be down-converted using a consistent phase reference and the analog-to-digital converters can take synchronous samples. This helps to ensure that relative phase information between the input signals is preserved in the digitized signals. In addition, the signal lines 526a-d from the local oscillator 528 to these signal components can be matched so as to further help maintain phase coherency in the receiver. Although
In some embodiments, the received signals are coherently received and analyzed. Phase information can be preserved between the various received signals. For example, the received signals can share a common local oscillator 528 used in the down-conversion processing and the signals can be synchronously sampled during digital conversion. Coherence at the receiver may entail synchronization of the signal channels in various forms, which can include: phase synchronization; frequency synchronization, sampling synchronization; and local oscillator synchronization in frequency, time, and/or phase. In some embodiments, the receiver 520 can also be coherent with the transmitter 510. For example, the transmitter 510 and the receiver 520 could share a common phase reference such as a local oscillator (e.g., as in a monostatic embodiment where the transmitter and receiver are housed together). (This can provide additional ways to characterize the transmitter-channel-receiver system by enabling, for example, the characterization of Doppler spreads induced in the system.) Additionally, it may be desirable that the receiver signal channels are gain and phase matched (from the antennas to the analog-to-digital converters) across all frequency components of interest and that the local oscillator signal gains to each channel are substantially matched. In some embodiments, the receiver 520 can advantageously achieve precise control of the phase, amplitude, sampling, and frequency among the various receiver channels.
As already mentioned, the receiver channels can be phase and/or gain matched. In some cases, the phase and/or gain matching can be dynamically adjusted. This can be accomplished using phase shifting elements and/or amplifiers in each receiver channel. In some embodiments, these phase shifting elements and/or amplifiers can be adjustable based on, for example, a calibration control input. The calibration control input can be obtained by passing a calibration signal through the various receiver processing channels. The effect of each processing channel on the calibration signal can then be determined. A calibration control input can be generated in order to reduce or eliminate differences between the effects that each processing channel has on the calibration signal. For example, a calibration control input can be generated in order to reduce or eliminate differences between the respective gains of the receiver channels and/or to reduce or eliminate phase differences between the channels. In addition, the phase and/or gain matching can be temperature compensated to help reduce phase and/or gain mismatches which may be induced at different operating temperatures. Digital compensation of the digitized signals can also be employed to achieve phase and/or gain matching.
Although the receiver 520 in
Once, the SR1uT1x, SR1vT1x, SR1uT1y, SR1vT1y, SR2uT1x, SR2vT1x, SR2uT1y, and SR2vT1y signals are down-converted and sampled, the respective frequency component phases and amplitudes for various signal pairs can be compared as a means of learning information about the transmitter-channel-receiver system. The different signal pairs are described below with respect to
The first six signal pairs in
Additional non-polarization information about the multipath channel can be obtained by also comparing respective frequency component phases and/or amplitudes from signals detected at different antennas. Signal pairs 3-6 in
The second six signal pairs in
The next 16 signal pairs in
The first 28 signal pairs in the table shown in
While
While separate transmitter and/or receiver signals have been described herein as being associated with the individual outputs of separate antenna ports, it is not required that each transmitted signal correspond only to what is sent via a single antenna or that each received signal correspond only to what is received via a single antenna. For example, instead of employing antenna ports as the fundamental quantity, beams derived from a weighted combination of antenna elements (on the transmitter and/or receiver side) can be used instead. In such cases, each beam can be treated as one of the transmitter/receiver signals for purposes of the analysis described herein. This is one of the benefits of a coherent system. In fact, these beams can even be frequency dependent. For a linear combination of spatially-separated antennas, frequency-dependent weights could correspond to different beam steering directions as a function of frequency. For linear combinations of a single dual polarized antenna, frequency-dependent weights would generally correspond to different polarizations as a function of frequency. For an antenna system with both space and polarization separated elements, a weighted combination involving space and polarization dimensions can be used.
While
In addition, although
At block 630, the transmitted and received signals from blocks 610 and 620 can each be separated into frequency sub-bands. This can be done using, for example, a Fourier transform or other processing.
At block 640, multiple pairs of received and transmitted signals are formed.
At block 650, frequency component phase and/or amplitude comparison data can be calculated for each signal pair from block 640 and for each frequency sub-band from block 630. For example, the amplitudes of the frequency components of one of the signals can be compared to those of the other by calculating differences between the respective amplitudes or ratios of the amplitudes. Similarly, the phases of the frequency components of one of the signals can be compared to those of the other by calculating differences between the respective phases. Other computations can also be useful in comparing these magnitudes and phases. For example, in some embodiments, calculation of the phase and/or amplitude comparison data is accomplished by calculating a Jones vector or Stokes parameters (normalized or unnormalized) for each sub-band of each signal pair. (Again Stokes parameters (S0, S1, S2, and S3) for each sub-band can be calculated according to the following equations: S0=(Y1·Y*1)+(Y2·Y*2); S1=(Y1·Y*1)−(Y2·Y*2); S2=(Y1·Y*2)+(Y2·Y*1); and S3=j(Y1·Y*2)−j(Y2·Y*1), where Y1 is a complex number with amplitude and/or phase information for a first signal in the pair of signals being compared and Y2 is a complex number with amplitude and/or phase information for a second signal in the pair of signals being compared.) Although these computations are traditionally used to determine polarization states, they can also be applied as an analytical tool even in cases where the signal pairs are such that the computations do not result in polarization information. As discussed herein, the set of per sub-band comparison values for each signal pair can be referred to as a coherent signal dispersion (CSD) curve or a polarization mode dispersion (PMD) curve, depending on the particular signal pair.
As just mentioned, for each signal pair obtained from any system architecture described herein, Jones vectors or Stokes vectors can be formed. The representation for the former can be written as a complex scale factor (amplitude and phase) that multiplies a unit Jones vector. If relative amplitude and relative phase alone are of interest (such as in characterizing polarization states on a unit sphere), the complex scale factor can be ignored, although the amplitude and phase information provided by the complex scale factor can potentially be useful for sensing and other applications. Stokes vectors of the form [S0 S1 S2 S3] can be formed for each signal pair using, for example, the equations provided herein. This unnormalized form of a Stokes vector may or may not have a degree of polarization of unity (i.e., where the square of S0 equals the sum of the squares of S1, S2, and S3). In some embodiments, however, the sub-band spacing can be chosen so that the degree of polarization is near unity. In some cases, it may be appropriate to normalize the [S1 S2 S3] vector (e.g., so that the sum of the squares of S1, S2, and S3 equals the square of S0, which essentially “forces” the condition of having unit degree of polarization). When plotting the CSD or PMD curves in any of these cases, the 3D locus will not be constrained to a unit sphere, but in some cases, it may useful to normalize the [S1 S2 S3] vectors to have unit magnitude so that the CSD or PMD curves will be constrained to a unit sphere. In the case of PMD, this is equivalent to considering the polarization state (i.e., the relative amplitude and relative phase between the signals associated with the signal pair). Since these representations deal primarily with relative amplitude and relative phase information, some amplitude and phase information (a complex scale factor) is not retained through this representation. For all of the cases, it may be useful to retain amplitude and/or phase information associated with the signal pairs that might otherwise be lost in a particular representation. The amplitude and phase can be relative to some reference used to measure these values.
Calculation of a set of Stokes parameters for each sub-band results in a Stokes vector for each sub-band. (Again, although the same equations may be used for calculating Stokes vectors for CSD signal pairs as for PMD signal pairs, the Stokes vectors for CSD signal pairs do not consist of polarization information). If the Stokes vectors (and hence the curves) are not normalized to unit magnitude, the vectors contain amplitude information (e.g., the S0 term in the Stokes vector provides amplitude information) that can be utilized in addition to phase information to analyze the signals. The resultant CSD (or PMD) curve from non-normalized Stokes vectors would not necessarily be constrained to reside on a unit sphere. In some cases, CSD and PMD curves may be continuous. However, in some cases, the resulting curve is a locus of points that may not be continuous. For example, if the transmit polarization is varied with sub-band, or more generally, if the relative amplitude and phase between transmit ports is varied with sub-band, the resulting curve may exhibit discontinuities.
For each signal pair, frequency component amplitude and/or phase comparisons can be made between the signals for different relative delays (e.g., where one of the signals is delayed by one or more samples), or for different frequency offsets (for example where the subcarriers of the two signals are not the same, but are intentionally offset). These offsets in delay and frequency can also be considered simultaneously (e.g., offsets in delay and in frequency). Such characterizations may be useful to establish decorrelation times and decorrelation frequencies. Furthermore, a signal pair consisting of a receiver signal and a transmitter signal could use a delay difference for the signals to align them in time for comparison purposes. Signal cross-correlation, for example, could be used to identify the delay that should be used to align the transmitter signal with the receiver signal.
Dynamic CSD curves can be determined by applying the just-described technique repeatedly over time. This can be done by extracting a time window of data of a desired length from the pairs of received/transmitted signals. Then, for each time window, the frequency component phase and/or amplitude comparison data can be calculated for each frequency sub-band. The time window can then be advanced and the per sub-band comparison values can be calculated once again. This process can be repeated as long as desired in order to determine the time domain behavior of the CSD curves. The length of the time window for each of these iterations can be selected, for example, based upon the timescale of the time-varying effects that are to be analyzed.
At block 660, the frequency component phase and/or amplitude comparison data (e.g., coherent signal dispersion (CSD) curves) from block 650 can be analyzed in order to determine a characteristic of the transmitter, receiver, and/or channel, including a characteristic of a target located in the channel. In some embodiments, this analysis can include visualization by plotting the per sub-band comparison data for each signal pair on or about a sphere or other manifold.
The analysis in block 660 can include identifying a characteristic of the comparison data from block 660 at a given time (e.g., length, shape, location on the sphere of a CSD curve, etc.). A characteristic of interest can be identified by, for example, relating the comparison data to calibration data or previously-elicited comparison data. Additionally, the analysis can include identifying a change in a characteristic of the comparison data as a function of time (e.g., length, shape, location on the sphere of a CSD curve, etc.). A characteristic of the comparison data may correspond to a physical characteristic of the system. For example, the length of a CSD curve may be reflective of temporal dispersion between channels; the complexity of a CSD curve may be indicative of the multipath composition; and periodic oscillations may reflect periodic processes in the transmitter-channel-receiver system. Any of these properties, or others, of the comparison data can be analyzed. These analyses can be conducted in the time domain, spatial domain, and/or frequency domain. For example, assume that a target within the channel vibrates at a frequency, fv, while the transmitter and receiver are held stationary. A spectral analysis, perhaps via a discrete Fourier transform, of one or more of the dynamic Stokes parameters calculated from PMD or CSD data should indicate the presence of a frequency component at fv. The magnitude of this fv component along with the possible presence of other frequency components could provide useful information about said vibrating target. Thus, the spectral analysis can include, for example, determining the magnitude(s) of one or more spectral components of the comparison data from block 660. Many techniques are disclosed in U.S. Patent Publication 2013/0332115 for analyzing polarization mode dispersion curves to obtain useful information about a multipath channel. Notwithstanding the distinctions between polarization mode dispersion curves and coherent signal dispersion curves, the same PMD curve analysis techniques can be applied to the CSD curves disclosed herein. Therefore, U.S. Patent Publication 2013/0332115 is incorporated by reference herein in its entirety for its disclosure of such analysis techniques.
Various operations that can be performed on the coherent signal dispersion curves as part of these analyses include filtering, averaging, statistical analyses, excision, integration, rotation, smoothing, correlation, eigendecomposition, Fourier analyses, and many others.
For some analyses it may be advantageous to reduce each coherent signal dispersion curve to a single value that represents the curve as a whole. This can be done using, for example, a centroiding operation. Experiments have shown that the centroid of a coherent signal dispersion curve can efficiently and effectively reduce unwanted noise while still providing useful information about the transmitter-channel-receiver system.
Estimation techniques can be applied in order to reduce variations in a measured CSD curve. This can be done because there typically is a correlation between the values for neighboring sub-bands in the curve (i.e., the coherence signal dispersion information is not generally expected to exhibit discontinuities from one sub-band to the next). This property of coherent signal dispersion curves allow for the usage of techniques to improve the quality of CSD curve estimates.
CSD curves are believed to be dependent to a significant degree on the transmitter-channel-receiver system, including the state of any targets within the channel. (The CSD curves may be dependent to a lesser degree—potentially a far lesser degree—on the specific content or properties of the transmitted signals, for example, so long as the transmitted signals have adequate signal strength across the bandwidth being analyzed.) In other words, the CSD curves are believed to be strongly dependent on the factors impacting the transmitter (such as transmit antenna location/motion, transmit polarization, beam pattern, etc), the receiver (such as receiver antenna location/motion and beam pattern), and factors leading to the channel response. The CSD curves will change in response to physical changes in the frequency-selective environment, including physical movement of scatterer targets in relation to the locations of transmitting and receiving antennas. This means that characteristics of the CSD curves at a given moment in time may be used to identify a specific multipath channel, including a specific state of a target located in the channel, potentially without knowledge of the transmitted signal(s) that produced the CSD curves.
One application of this property is that the transmitted signal(s) need not necessarily be known in order to determine useful information about a target located in the channel. Instead, a signal of opportunity can be used as the transmitted signal. Signals of opportunity could include, for example, cellular telephone signals, Wi-Fi signals from an Internet hotspot, and many others. These signals can be received and analyzed using the systems and techniques discussed herein to learn information about, for example, a target located in the environment. One specific application which could entail the use of a signal of opportunity is a system for measuring a patient's heart or respiration rate in a hospital or other clinical environment. Such environments typically have strict regulations regarding the transmission of wireless signals. Thus, it could be advantageous if the system did not require its own transmitter but could instead make use of unknown existing signals of opportunity. The system could generate one or more CSD curves by receiving and processing those existing transmitted signals, as discussed herein. If the patient's heart or lungs are present in the propagation channel between the receiver and the unknown transmitted signals of opportunity, then one or more of the CSD curves will likely include information about the rate of movement of the heart or lungs. This rate of movement can be determined by, for example, analyzing the frequency content of the CSD information.
Another application of the CSD analysis described herein relates to monitoring the movements of, for example, mechanical machinery. In the case of fixed transmit and receive antennas, such movements, even if they are small vibrations, can result in changes to the multipath wireless environment of the object. As already noted, these changes in the multipath environment can lead to corresponding changes to the CSD curves that are detected using the systems and methods described herein. Changes in the CSD curves can be analyzed in order to monitor the normal operation of the machinery or even detect irregular operation, such as new or different vibrations. Take the example of a three-blade fan. The rotational frequency of the fan can be determined from the CSD curves because they will vary at a rate that corresponds to the rotational frequency of the fan. Further, if a ball bearing begins to fail, or one of the fan blades becomes damaged, this will induce a change in the vibrations that can also be detected by monitoring changes in the CSD curves. Many techniques are disclosed in U.S. Patent Publication 2013/0332115 for analyzing polarization mode dispersion curves to obtain useful information about such physical movements of a target object. Notwithstanding the distinctions between polarization mode dispersion curves and coherent signal dispersion curves, the same PMD curve analysis techniques can be applied to the CSD curves disclosed herein. Therefore, U.S. Patent Publication 2013/0332115 is incorporated by reference herein in its entirety for its disclosure of such analysis techniques.
One benefit of the CSD curves described herein over the PMD curves described in U.S. Patent Publication 2013/0332115 is the rich diversity of the CSD curves, which far outnumber PMD curves. Owing to the rich diversity of the CSD curves, it becomes much more likely that a given time-varying characteristic of the multipath channel, including a target object in the channel, will be evident in at least one of the CSD curves.
U.S. Patent Publication 2013/0332115 describes many other practical applications of PMD analysis. It should be understood that the systems and methods described herein for performing CSD can also be applied to any of those applications, likely with improved results. Thus, U.S. Patent Publication 2013/0332115 is incorporated by reference herein for its disclosure of all such practical applications.
Any of the systems and methods described herein can be used to obtain coherent signal dispersion (CSD) information in order to monitor rotating machinery, such as turbomachinery. This can be done by providing transmitter and receiver antenna probes with access to the internal cavities of the rotating machinery. These probes can be respectively connected to any of the transmitters and receivers described herein in order to obtain signals which can be analyzed to learn information about the rotating machinery. The systems and methods described herein can be used for detection of a wide variety of physical phenomenon within a rotating machine.
It is possible to use electromagnetic signals to monitor rotating machinery because the dielectric properties of metals impact electromagnetic signals in, for example, the gigahertz (GHz) range. Hence, RF signals propagating throughout the internal cavities of a rotating machine will be affected by (e.g., modulated due to reflection, refraction, scattering, etc.) the physical changes (e.g., movements or vibrations) of the metal components and boundaries comprising the transmitter-to-receiver propagation channel. These physical movements and/or other changes affect the structure of the multipath channel inside the rotating machinery and can result in, for example, time-varying temporal multipath dispersion properties that manifest themselves as dispersion signatures over the bandwidth of the interrogating RF signal.
The signals transmitted and received by antenna probes inside rotating machinery can include features induced by periodic rotations of the machinery (e.g., motion of a rotor, turbine or compressor vanes/blades, etc.) and by anomalous events (e.g., shaft unbalance, bearing fatigue, blade deformation/vibration, the onset of stalls, etc.). These features in the signals can be analyzed to determine information about the machinery. For example, the information collected from rotating machinery using the systems and methods described herein can be used to detect or otherwise identify, in real-time, precursors of undesired occurrences, including stalls, surges, and catastrophic failures. By detecting or otherwise identifying precursors of these events, action can be taken (e.g., control inputs can be modified) to prevent such events or to reduce their severity. Analysis of the signals from the antenna probes can also be used to improve the design of machinery, identify manufacturing defects, or to provide diagnostic information during prototype or characterization phases. The systems and methods described herein can therefore make it possible to improve the design/development process and to operate turbomachinery (or other types of rotating machinery) with higher efficiencies, lower costs, and reduced maintenance downtime.
The example gas turbine engine in
As shown in
Finally,
As discussed further with respect to
Once coherent signal dispersion (CSD) data has been obtained from the rotating machinery using antenna probes connected to the transmitters and receivers described herein, the CSD data can be analyzed using techniques also described herein. For example, monitoring schemes can be established based on inter-signal correlations and relative amplitude and relative phase of transfer functions between different transmitter and/or receiver signal pairs.
This can be done by first forming signal pairs between various transmitter and/or receiver signals as discussed herein. Monitoring all possible pairwise combinations of signals provides immense diversity to increase the probability of detecting even small changes in the performance or operation of the rotating machinery. The signals available from the system architectures described herein are rich in information content, possessing joint correlation properties in space, polarization, etc. that can be leveraged for sensing.
Each signal can be divided into multiple frequency sub-bands. Dividing the full bandwidth signal into smaller sub-bands can improve coherence properties and improve signal characterizations. In addition, the various sub-bands provide added diversity in characterizing the RF signal, and therefore in measuring changes that result from multipath changes induced by shaft unbalances, blade deformation, etc.
Amplitude and/or phase information can then be determined for each sub-band. Then the amplitude and/or phase information for one signal in each pair can be compared to the corresponding amplitude and/or phase information for the other signal in the pair. The resulting comparison data can take several forms. For example, Stokes parameters, or the like, can be calculated for each sub-band of each signal pair.
The Stokes parameters (or other comparison data) can then be analyzed using a number of signal processing techniques. In some embodiments, the Stokes parameters (or other comparison data) are analyzed on a per sub-band basis. In other embodiments, the Stokes parameters (or other comparison data) from multiple sub-bands can be combined by performing a centroiding operation. In either case, a time series of Stokes parameters or centroid data can be analyzed using Fourier analysis or other similar frequency domain analysis techniques. These techniques can be used, for example, to identify one or more frequency components or to identify changes in the frequency content of the signals over time. Time domain processing can also be performed to identify or analyze signal features of interest. Many other signal processing techniques can also be used.
U.S. Patent Publication 2013/0332115 describes systems and methods for obtaining and analyzing polarization mode dispersion (PMD) information from rotating machinery. As already mentioned, the systems described herein can be used to obtain coherent signal dispersion (CSD) information from rotating machinery. Nevertheless, the same analysis techniques can be applied to the CSD information as are disclosed in U.S. Patent Publication 2013/0332115 with respect to PMD information. U.S. Patent Publication 2013/0332115 is therefore incorporated by reference herein for its disclosure of such analysis techniques.
If the data in
By continually monitoring the CSD data, any slight change in given characteristic of the data can be used as an indicator, or even a future predictor, of a defect, fault, or failure of the rotating machine. The CSD data can be used in a feedback control system to prevent or reduce the severity of an undesired operating condition. For example, if a predictor of a defect, fault, or failure is identified (e.g., using real-time processing), then the control system can alter a control input (e.g., reduce power, etc.) in an effort to prevent the defect, fault, or failure from occurring. Although a future predictor of a defect, fault, or failure may be due to an internal cause, the systems and methods described herein can also detect external causes. For example, antennas can be mounted at different axial and/or radial positions about the air intake of a gas turbine engine. These antennas can be used to generate and capture signals which can be analyzed to detect foreign matter (e.g., birds, etc.) either before or just as it enters the air intake. In response to detection of foreign matter, the engine can be shut down or otherwise controlled so as to reduce damage to the engine from such foreign matter.
Embodiments have been described in connection with the accompanying drawings. However, it should be understood that the figures are not drawn to scale. Distances, angles, etc. are merely illustrative and do not necessarily bear an exact relationship to actual dimensions and layout of the devices illustrated. In addition, the foregoing embodiments have been described at a level of detail to allow one of ordinary skill in the art to make and use the devices, systems, etc. described herein. A wide variety of variation is possible. Components, elements, and/or steps may be altered, added, removed, or rearranged. While certain embodiments have been explicitly described, other embodiments will become apparent to those of ordinary skill in the art based on this disclosure.
The systems and methods described herein can advantageously be implemented using, for example, computer software, hardware, firmware, or any combination of software, hardware, and firmware. Software modules can comprise computer executable code for performing the functions described herein. In some embodiments, computer-executable code is executed by one or more general purpose computers. However, a skilled artisan will appreciate, in light of this disclosure, that any module that can be implemented using software to be executed on a general purpose computer can also be implemented using a different combination of hardware, software, or firmware. For example, such a module can be implemented completely in hardware using a combination of integrated circuits. Alternatively or additionally, such a module can be implemented completely or partially using specialized computers designed to perform the particular functions described herein rather than by general purpose computers. In addition, where methods are described that are, or could be, at least in part carried out by computer software, it should be understood that such methods can be provided on computer-readable media (e.g., optical disks such as CDs or DVDs, hard disk drives, flash memories, diskettes, or the like) that, when read by a computer or other processing device, cause it to carry out the method.
A skilled artisan will also appreciate, in light of this disclosure, that multiple distributed computing devices can be substituted for any one computing device illustrated herein. In such distributed embodiments, the functions of the one computing device are distributed such that some functions are performed on each of the distributed computing devices.
While certain embodiments have been explicitly described, other embodiments will become apparent to those of ordinary skill in the art based on this disclosure. Therefore, the scope of the invention is intended to be defined by reference to the claims and not simply with regard to the explicitly described embodiments.
This invention was made with government support under contract N00014-12-1-0539 awarded by the U.S. Office of Naval Research and under contract 2011-11070800002 from the Central Intelligence Agency.
| Number | Name | Date | Kind |
|---|---|---|---|
| 4907943 | Kelch | Mar 1990 | A |
| 5235340 | Shea | Aug 1993 | A |
| 5291322 | Itoh et al. | Mar 1994 | A |
| 6324030 | Cheung et al. | Nov 2001 | B1 |
| 6845243 | Gaddy et al. | Jan 2005 | B1 |
| 9325436 | Pratt et al. | Apr 2016 | B2 |
| 20030092456 | Dent | May 2003 | A1 |
| 20030095607 | Huang et al. | May 2003 | A1 |
| 20030153273 | Ebert et al. | Aug 2003 | A1 |
| 20040077379 | Smith et al. | Apr 2004 | A1 |
| 20050281343 | Hsu et al. | Dec 2005 | A1 |
| 20070129025 | Vasa et al. | Jun 2007 | A1 |
| 20070253461 | Billington | Nov 2007 | A1 |
| 20080019263 | Stolpman | Jan 2008 | A1 |
| 20080045832 | McGrath | Feb 2008 | A1 |
| 20080074307 | Boric-Lubecke et al. | Mar 2008 | A1 |
| 20090078052 | Twerdochlib | Mar 2009 | A1 |
| 20090203972 | Heneghan et al. | Aug 2009 | A1 |
| 20100003034 | Pratt | Jan 2010 | A1 |
| 20100143963 | Pollack et al. | Jun 2010 | A1 |
| 20100150013 | Hara et al. | Jun 2010 | A1 |
| 20100256462 | Rappaport et al. | Oct 2010 | A1 |
| 20100292559 | Hannemann et al. | Nov 2010 | A1 |
| 20110274442 | Zhang et al. | Nov 2011 | A1 |
| 20120069355 | Hynous | Mar 2012 | A1 |
| 20130332115 | Pratt | Dec 2013 | A1 |
| 20140253368 | Holder | Sep 2014 | A1 |
| 20150222022 | Kundtz et al. | Aug 2015 | A1 |
| 20150226709 | Russhard | Aug 2015 | A1 |
| 20160015352 | Brown et al. | Jan 2016 | A1 |
| 20170131335 | Pratt et al. | May 2017 | A1 |
| Number | Date | Country |
|---|---|---|
| WO 2017136012 | Aug 2017 | WO |
| WO 2018145042 | Aug 2018 | WO |
| Entry |
|---|
| Pratt, T., “A Modified XPC Characterization for Polarimetric Channels”, Sep. 2011, IEEE Transactions on Vehicular Technology, vol. 60, p. 2904-2913. |
| International Search Report and Written Opinion received in PCT Application No. PCT/US2016/060821 dated Jul. 13, 2017 in 9 pages. |
| International Search Report and Written Opinion received in PCT Application No. PCT/US2018/016927 dated May 11, 2018 in 11 pages. |
| Number | Date | Country | |
|---|---|---|---|
| 20170134154 A1 | May 2017 | US |
