State filter

Inventors

Olson, John V.

Assignees

University of Alaska Fairbanks

Publication Number

US-8554816-B2

Publication Date

2013-10-08

Expiration Date

2030-08-26

Interested in licensing this patent?

MTEC can help explore whether this patent might be available for licensing for your application.


Abstract

Embodiments described herein describe the construction of frequency domain estimates of generalized power density and the filters that can be constructed from those estimates. Using the concept of the Stokes vector representation of the spectral matrix in an M-dimensional vector space, a generalization of the process in which the spectral matrix may be represented by a set of trace-orthogonal matrices that are based upon a particular signal state can be produced. One aspect of the process is as follows: given a particular signal, represented as by a state vector in the space, a complete, orthonormal set of vectors can be produced that includes the signal of interest. Then, a generalized set of matrices is constructed, based upon the developed vectors, that are trace-orthogonal and which serve as a basis set for the expansion of the spectral matrix. The coefficients of this expansion form a generalized Stokes vector that represents the power in the spectral matrix associated with the various state vectors. Filters that serve to extract or suppress information about a particular state may then be constructed using the components of the generalized Stokes vectors. The effectiveness of an embodiment this filter is demonstrated using acoustic data from a microphone array.

Core Innovation

The invention described provides systems, methods, and computer program products for a state filter that processes multivariate data signals to extract or suppress information about particular signal states. Utilizing the Stokes vector representation of the spectral matrix in an M-dimensional vector space, the method constructs frequency domain estimates of generalized power density. The spectral matrix is expanded using a set of trace-orthogonal matrices based upon an orthonormal set of vectors that includes the signal state of interest.

The problem addressed is the need to resolve the 'cocktail party effect'—the ability to focus attention on a single source in a mixture of signals and background noise—by eliminating dominant background noise and enabling the perception of quieter signals. The invention provides a mechanism by which a filter can be constructed using the components of generalized Stokes vectors to selectively extract or suppress information related to particular signal states, such as specific sources of noise or clutter, in a multivariate data environment.

The process involves receiving multivariate data and information about a particular signal state, generating a set of orthonormal vectors containing the state vector of interest, and constructing trace-orthogonal matrices from these vectors as a basis for spectral matrix expansion. The coefficients from this expansion form the generalized Stokes vector, representing the power in the spectral matrix associated with each state. Filters are then constructed using these coefficients to extract or suppress information about the desired signal state. The effectiveness of the filter is demonstrated using acoustic data from a microphone array, illustrating the suppression of dominant signals and recovery of weaker signals at the same frequency.

Claims Coverage

The patent contains independent claims that detail several inventive features relating to the construction and application of state filters using the generalized Stokes vector representation in multivariate signal processing.

Computer-implemented state filter method using generalized Stokes vectors

The invention comprises a computer-implemented method for filtering multivariate data by: 1. Receiving a multivariate data set and information about a particular signal state represented by a state vector of interest. 2. Generating a set of orthonormal vectors from the multivariate data set containing the state vector of interest. 3. Constructing a generalized set of trace-orthogonal matrices based on these vectors, serving as a basis set for spectral matrix expansion. 4. Forming a generalized Stokes vector from coefficients of the expansion, representing power in the spectral matrix associated with the state vector of interest. 5. Constructing filters that extract or suppress information about a particular state using components of the generalized Stokes vector.

Device for filtering using generalized Stokes vectors

A device comprising a memory and a processor configured to execute computer-executable code sections for: - Receiving a multivariate data set. - Receiving information about a particular signal state represented by a state vector of interest. - Generating a set of orthonormal vectors from the multivariate data set that contains the state vector of interest. - Constructing a generalized set of trace-orthogonal matrices based on these vectors, as a spectral matrix expansion basis set. - Forming a generalized Stokes vector from the expansion coefficients, representing power associated with the state vector of interest. - Constructing filters to extract or suppress information about a particular state using components of the generalized Stokes vector.

System with sensors and processor for state-specific filtering in multivariate data

A system comprising: - A memory and a processor. - One or more sensors operably connected with the processor and memory. - The processor configured to execute code sections for: - Receiving a multivariate data set, obtained at least in part from sensors. - Receiving information about a particular signal state, represented by a state vector of interest. - Generating a set of orthonormal vectors containing the state vector of interest. - Constructing a generalized set of trace-orthogonal matrices as a basis for the spectral matrix expansion. - Forming a generalized Stokes vector from expansion coefficients for representing power in the spectral matrix. - Constructing filters using components of the generalized Stokes vector to extract or suppress information about a particular state.

Computer program product enabling generalized Stokes vector based filtering

A computer-executable program product embodied on a computer-readable medium for: - Receiving a multivariate data set. - Receiving information about a particular signal state, represented by a state vector of interest. - Generating a set of orthonormal vectors containing the state vector of interest. - Constructing a generalized set of trace-orthogonal matrices for expanding the spectral matrix. - Forming a generalized Stokes vector from expansion coefficients corresponding to power in the spectral matrix. - Constructing filters to extract or suppress information about a particular state using the generalized Stokes vector components.

These inventive features collectively define a framework—implemented as methods, devices, systems, and computer program products—for filtering multivariate data sets. The filtering leverages generalized Stokes vector expansion enabled by trace-orthogonal matrices tailored to a signal state of interest, supporting targeted extraction or suppression of specific information in the frequency domain.

Stated Advantages

Eliminates dominant background noise in multivariate data and allows quieter signals underneath to be perceived.

Enables suppression of unwanted signal states by as much as 20 dB or more when the spectral matrix represents a pure state.

Allows minor signals to survive the filtering process unless their state vector matches that of the suppressed signal.

Provides a general approach applicable to any multivariate data, including from arrays of like or different sensors.

The inner product in the Stokes vector space has operational significance and allows for a generalized power spectrum interpretation.

Facilitates the discrimination between signal and clutter based on differences in signal state vectors, supporting desirable signal extraction.

Documented Applications

Suppression of dominant, coherent signals and clutter reduction in acoustic data recorded by an infrasound microphone array, as demonstrated with data from the International Monitoring System (IMS) of the Comprehensive Nuclear Test-Ban Treaty Organization (CTBTO).

Extraction and recovery of weak signals in multivariate data that are otherwise obscured by stronger, unwanted signals.

JOIN OUR MAILING LIST

Stay Connected with MTEC

Keep up with active and upcoming solicitations, MTEC news and other valuable information.