Continuously adaptive dynamic signal separation and recovery system

• US 6,236,862 B1
• Filed: 12/15/1997
• Issued: 05/22/2001
• Est. Priority Date: 12/16/1996
• Status: Expired due to Fees

• Current Assignee: Innovative Computing Technologies, Inc.
• Original Assignee: Intersignal, LLC
• Inventors: Salam, Faihi M., Erten, Gamze
• Primary Examiner(s): Trost, William G.
• Assistant Examiner(s): SMITH, SHEILA B
• Application Number:
  US08/990,682
• Time in Patent Office: 1,254 Days
• Field of Search: 455/101, 455/102, 455/103, 455/105, 455/227, 455/228, 455/229, 455/280, 455/561, 455/501, 455/506, 370/320, 370/335, 370/342
• US Class Current: 455/501
• CPC Class Codes: G01S 13/006 Theoretical aspects G01S7/418, G01S13/9094, G01S13/958 take precedence
• CPC Class Codes: G01S 3/00 Direction-finders for determining the direction from which infrasonic, sonic, ultrasonic, or electromagnetic waves, or particle emission, not having a directional significance, are being received position-fixing by co-ordinating a plurality of determinations of direction or position lines G01S5/00
• CPC Class Codes: G01S 7/00 Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
• CPC Class Codes: G01S 7/52023 Details of receivers
• CPC Class Codes: H01Q 3/2605 Array of radiating elements provided with a feedback control over the element weights, e.g. adaptive arrays
• CPC Class Codes: H04B 1/123 using adaptive balancing or compensation means adaptive filter circuits and algorithms H03H
• CPC Class Codes: H04W 16/14 Spectrum sharing arrangements between different networks

Abstract

A method and apparatus for dynamically separating and recovering original signal sources by processing a set of mixed received mixtures and convolution of said signals utilizing differential equations and a computer. The system of the invention enables the blind separation and recovery of an unknown number of signals mixed together in dynamically changing interference environments with very minimal assumption on the original signals. The system of this invention has practical applications to nonmultiplexed media sharing, adaptive interferer rejection, acoustic sensors, acoustic diagnostics, medical diagnostics and instrumentation, speech, voice, language recognition and processing, wired and wireless modulated communication signal receivers, and cellular communications.

Claims

1. A signal processing system for separating signals received from plural sources via an unknown channel or medium, comprising a plurality of detectors each for detecting said mixed signals received from said plural sources and a processor for receiving the detected signals and processing them in accordance with the following compact general dynamic feedback vector equation:

2. A signal processing system according to claim 1, where entries of T and the values of η

_ij are sufficiently small or decaying, and Y and D are initialized by arbitrary values, and some entries of D are assigned to constant values and the update equation in claim 1 applies only to the entries of D which are not constant.

3. A signal processing system according to claims 1 or 2, wherein the number of signal sources that are detected by the detectors is equal to a nonnegative integer value N and the value of N is known, and the number of detector outputs is equal to the number m, and m is known, and N=m, in which case Y and E defined in claim 1 have the same dimension, D is a square matrix and the equations defined in claim 1 can be used directly without alteration.

4. A signal processing system according to claims 1 or 2, wherein the number of signal sources that are detected by the detectors is equal to a nonnegative integer value N and the value of N is known, and the number of detector outputs is equal to the number m, and m is known, and N≦

m, in which case the equations defined in claim 1 can be used directly without alteration, but the vector Y, which is of size m (r=m) and whose entries approximate the original sources will have some zero entries.

5. A signal processing system according to claims 1 or 2, wherein the number of signal sources that are detected by the detectors is equal to a nonnegative integer value N and the value of N is known, and the number of detector outputs is equal to the number m, and m is known, and N≦

m, in which case the equations defined in claim 1 can be used directly without alteration, and where the matrix B has less rows than columns, and where the vector Y is of size N (r=N) and the vector BY is of size N.

6. A signal processing system according to claims 1 or 2, wherein the number of signal sources that are detected by the detectors is equal to a nonnegative integer value N and the value of N is known, and the number of detector outputs is equal to the number m, and m is known, and N≦

m, in which case the equations defined in claim 1 can be used directly without alteration, and where the matrix B has less rows than columns, and where the vector Y is of size q (r=q) and the vector BY is of size q, such that N<

q<

m.

7. A signal processing system according to claims 1 or 2, wherein the number of signal sources that are detected by the detectors is equal to a nonnegative integer value N and the value of N is known, and the number of detector outputs is equal to the number m, and m is known, and N≧

m, in which case the equations defined in claim 1 are applied after preprocessing of detected mixed signals to eliminate or filter out a number of signal sources by other known methods in such a way that the remaining signal sources in the mixtures are less than or equal to the number of detectors.

8. A signal processing system according to claims 1 or 2, wherein the number of signal sources that are detected by the detectors is equal to a nonnegative integer value N and the value of N is known, and the number of detector outputs is equal to the number m, and m is known, and N≧

m, in which case the equations defined in claim 1 are applied nonetheless, after which a post processing step removes the signal sources that contaminate the output channels by other known methods.

9. A signal processing system according to claims 1 or 2, wherein the number of signal sources that are detected by the detectors is equal to a nonnegative integer value N and the value of N is known, and the number of detector outputs is equal to the number m, and m is known, and N≧

m, in which case the equations defined in claim 1 are applied nonetheless, but where the initialization values of D and Y can be manipulated in such a manner to direct one or more of the relevant or desired signals into separate channels while the rest of the “

undesired”

or irrelevant signals are directed to unused or discarded separation outputs.

10. A signal processing system according to claims 1 or 2, wherein the number of signal sources that are detected by the detectors is equal to m, and m is unknown, in which case their number is initially assumed to be equal to that of the number of detectors, which is equal to N, and based on the characteristics and contents of the output.

11. A signal processing system as defined in claim 1 wherein the physical computing mechanism is an architecture of analog or digital or mixed analog digital circuits, or other physical means that compute a set of elementary or complex functions arranged specifically for obtaining the solution of the differential equations in claim 1.

12. A signal processing system as defined in claim 1 wherein the physical computing mechanism is any general computer architecture which can be programmed with the necessary set of operations needed for obtaining the solution of the differential equations in claim 1.

13. A signal processing system as defined in claim 1 wherein the physical computing mechanism is an architecture specifically designed for real time on line instantaneous signal processing containing circuit elements that convert analog detector output to digital format, a digital signal processor programmed to obtain the solution of the differential equations in claim 1, and circuit elements for storage of data or programs necessary or desirable for such computation and elements that convert the digital output from the digital signal processor output back to analog form.

14. A channel sharing method where (1) a number of channels need to be shared among multiple transmitter sources, which may operate independently of one another, and (2) the number of said channels is less than the number of transmitter sources, and (3) the multiple transmitter sources each transmit messages using any set of modulation techniques, which may vary among transmitters, and some of which may be unknown, and (4) the said multiple transmitter sources transmit messages in a channel simultaneously by way of modulating the messages onto several asynchronous carriers, thus generating transmitter signals that may have overlapping spectra, and (5) such transmitted signals are mixed with one another in the channel subject to the physical properties of the channel so that the resulting composite signal in the channel may consist of the superposition of the transmitter signals, and (6) at the receiver end, a receiver, which uses multiple antennae or mimics the use of multiple antennae, detects or constructs simultaneous multiple copies of the superimposed transmitter signals with at least one received or reconstructed copy being different from the other copies and (7) the said multiple copies are processed to obtain the original transmitted signals by a signal separation method or apparatus that separates the superimposed signals at least partially into their unmixed components and estimates at least one of the original transmitter signals or messages.

15. A channel sharing method as in claim 14 where the transmitter signals are mixed not deliberately but unintentionally and the method of claim 14 is used for purposes of interferer rejection or cancellation.

16. A channel sharing method as in claim 14 used in conjunction with another channel sharing method for purposes of recovering at least one transmitted signal, or distinguishing one transmitted signal from another, or interferer rejection or noise cancellation.

17. A channel sharing method as in claim 14 used in conjunction with various signal localization or spatial filtering techniques.

21. A method as described in claim 14 where the transmitter signals are mixed deliberately or unintentionally and where the transmitter signals have been generated by using codes which are not orthogonal, or where transmitter signals have been generated with codes unknown or not unique to a single transmitter, and the method in claim 14 is used for purposes of distinguishing one transmitted signal from another, or interferer rejection, or noise cancellation.

18. A channel sharing method where (1) a number of channels need to be shared among multiple transmitter sources, and (2) the number of said channels are less than the number of transmitter sources, and (3) the multiple transmitter sources transmit messages using one or more types of modulation techniques independent of one another in a narrowband or broadband channel simultaneously by way of modulating the signals onto several asynchronous carriers, which, prior to or after modulation, may have overlapping spectra, and (4) such transmitted signals are mixed with another in the channel so that the resulting composite signal in the channel may consist of the superposition of the modulated signals comprising the steps of receiving the transmitted signals, detecting multiple dissimilar copies of the superimposed transmission signals and processing by signal separation the dissimilar copies of the superimposed transmission signals to obtain the original transmitted signals in accord with the following equations:

19. A channel sharing method as in claim 18 where the transmitter signals are mixed not deliberately but unintentionally and the method is used for purposes of interferer rejection or cancellation.

20. A channel sharing method as in claim 18 where the transmitter signals are mixed deliberately as in the case of code division multiple access techniques, and the method in claim 18 is used for purposes of interferer rejection or noise cancellation.

22. A channel sharing method where (1) a single frequency band is to be shared among two transmitters, and (2) there is only a single frequency band available to allocate between the two transmitters, and (3) the two transmitter sources transmitting messages use channel sharing or modulation techniques, selectively using or not using any individual code or set of unique codes that would render the modulated signal distinguishable from other modulated signals that use different and distinguishable code, and (4) the two transmitter sources transmitting messages in the same frequency band simultaneously by way of modulating the signals onto several asynchronous carriers, which may have overlapping spectra, and (5) the two transmitted signals mix with each another in the same frequency band so that the resulting composite signal in the band consists of a superposition of the modulated signals, and (6) at the receiver end, a composite receiver detects two dissimilar copies of the superimposed transmission signals, so that one copy is different from the other copies, and (7) the two copies are processed to obtain the two original transmitted signals by a signal separation method.