Programmable methods of performing complex optical computations using data processing system

Claim 1

A method of programmably performing optical computation using a plurality of spatial light modulators to impress images corresponding to data onto a data beam, said method comprising the steps of: (a) programmably selectively performing a desired optical computation from a plurality of performable optical computations including Fourier transform, correlation, convolution and ambiguity function calculations; (b) applying respective data to a selected subset of said modulators to impress corresponding data images onto said beam; and (c) applying uniform data to the remaining ones of said modulators so as to impress corresponding uniform data images onto said beam such that said selected subset of said modulators is determined by the programmably selected optical computation.

Claim 2

A method of programmably performing optical computation using a plurality of spatial light modulators to impress images corresponding to data onto a data beam, said method comprising the steps of: (a) programmably selectively performing a desired optical computation from a plurality of performable optical computations including Fourier transform, correlation, convolution and ambiguity function calculations; (b) applying respective data to a selected subset of said modulators to impress corresponding data images onto said beam; and (c) applying uniform data to the remaining ones of said modulators so as to impress corresponding uniform data images onto said beam such that said selected subset of said modulators is determined by the programmably selected optical computation.

Claim 6

A method of programmably performing a cross correlation or sliding window cross correlation calculation of two-dimensional data with one- or two-dimensional reference data using an optical processor capable of impressing a plurality of data images onto a data beam using spatial light modulators, including crossed one-dimensional spatial light modulators, said method comprising the steps of: (a) buffering the one or two-dimensional data reference; (b) buffering the two-dimensional data matrix to be acted upon; (c) sequentially providing first data from the data reference to a first one-dimensional spatial light modulator so as to impress a first data image onto said beam; (d) sequentially providing second data from the two-dimensional data matrix to a second one-dimensional, spatial light modulator so as to impress a second data image onto said beam; and (e) spatially summing the data represented by the combination of said first and second data images on said data beam with a sequential provision of said first and second data.

Claim 7

A method of programmably performing a cross correlation or sliding window cross correlation calculation of two-dimensional data with one- or two-dimensional reference data using an optical processor capable of impressing a plurality of data images onto a data beam using spatial light modulators, including crossed one-dimensional spatial light modulators, said method comprising the steps of: (a) buffering the one or two-dimensional data reference; (b) buffering the two-dimensional data matrix to be acted upon; (c) sequentially providing first data from the data reference to a first one-dimensional spatial light modulator so as to impress a first data image onto said beam; (d) sequentially providing second data from the two-dimensional data matrix to a second one-dimensional, spatial light modulator so as to impress a second data image onto said beam; and (e) spatially summing the data represented by the combination of said first and second data images on said data beam with a sequential provision of said first and second data.

Claim 16

A method of performing a two-dimensional Fourier transform of a two-dimensional data matrix having rows and columns, on a data beam using a plurality of spatial light modulators and a spatial accumulator, said method comprising the steps of: (a) buffering a Fourier transform coefficient matrix having rows and columns for the first data dimension; (b) buffering said two-dimensional data matrix to be transformed; (c) applying the buffered Fourier transform coefficients of the last column of said Fourier transform coefficient matrix to the respective rows of a first one-dimensional spatial light modulator; (d) applying the buffered data of the last row of said two-dimensional data matrix to be transformed to the respective columns of a second one-dimensional spatial light modulator; (e) providing uniform data to all remaining ones of said spatial light modulators; (f) summing into said spatial accumulator the products of the intersecting data provided on said first and second one-dimensional spatial light modulators as effectively impressed on, combined, and received from said data beam; (g) respectively recirculating said two-dimensional data matrix and said Fourier transform coefficient matrix so as to provide new last columns and rows of respective coefficients and data to be applied to said first and second one-dimensional spatial light modulators; (h) repeating steps (c) through (g) for each row of said two-dimensional data matrix; (i) buffering said summed first data dimension resultant data from said spatial accumulator in place of said two-dimensional data matrix to be transformed; (j) buffering a Fourier transform coefficient matrix for the second data dimension in place of said Fourier transform coefficient matrix for the first data dimension; and (k) repeating steps (c) through (g) for each row of said summed first data dimension resultant data.

Claim 17

A method of performing a two-dimensional Fourier transform of a two-dimensional data matrix having rows and columns, on a data beam using a plurality of spatial light modulators and a spatial accumulator, said method comprising the steps of: (a) buffering a Fourier transform coefficient matrix having rows and columns for the first data dimension; (b) buffering said two-dimensional data matrix to be transformed; (c) applying the buffered Fourier transform coefficients of the last column of said Fourier transform coefficient matrix to the respective rows of a first one-dimensional spatial light modulator; (d) applying the buffered data of the last row of said two-dimensional data matrix to be transformed to the respective columns of a second one-dimensional spatial light modulator; (e) providing uniform data to all remaining ones of said spatial light modulators; (f) summing into said spatial accumulator the products of the intersecting data provided on said first and second one-dimensional spatial light modulators as effectively impressed on, combined, and received from said data beam; (g) respectively recirculating said two-dimensional data matrix and said Fourier transform coefficient matrix so as to provide new last columns and rows of respective coefficients and data to be applied to said first and second one-dimensional spatial light modulators; (h) repeating steps (c) through (g) for each row of said two-dimensional data matrix; (i) buffering said summed first data dimension resultant data from said spatial accumulator in place of said two-dimensional data matrix to be transformed; (j) buffering a Fourier transform coefficient matrix for the second data dimension in place of said Fourier transform coefficient matrix for the first data dimension; and (k) repeating steps (c) through (g) for each row of said summed first data dimension resultant data.

Claim 18

A method of performing a two-dimension cross correlation of a two-dimensional data matrix on a data beam using a plurality of spatial light modulators and a spatial accumulator, said method comprising the steps of: (a) buffering a two-dimensional correlation mask data matrix; (b) buffering said two-dimensional data matrix to be correlated; (c) applying the buffered said correlation mask data matrix to the corresponding spatial positions of a two-dimensional spatial light modulator; (d) applying a datum of said two-dimensional data matrix to be correlated, obtained from a predetermined row and column buffer position, to a zero-dimensional spatial light modulator; (e) providing uniform data to all remaining ones of said spatial light modulators; (f) summing into said spatial accumulator the products of the intersecting said correlation mask data matrix and said datum of said two-dimensional data matrix to be correlated, as provided to said two- and said zero-dimensional spatial light modulators, respectively, to thereby produce cumulative two-dimensional resultant data; (g) recirculating the buffered said two-dimensional data matrix to be correlated in a first data matrix direction so as to provide a new datum at said predetermined row and column buffer position; (h) recirculating said cumulative two-dimensional resultant data similarly in said first data matrix direction; (i) repeating steps (c) through (h) for each row or column in said first data matrix direction of said two-dimensional data matrix to be correlated; (j) recirculating the buffered said two-dimensional data matrix to be correlated in a second data matrix direction so as to provide a new datum at said predetermined row and column buffer position; (k) recirculating said cumulative two-dimensional resultant data similarly in said second data matrix direction; and (l) repeating steps (i) through (k) for each row or column in said second data matrix direction of said two-dimensional data matrix to be correlated.

Claim 19

A method of performing a two-dimension cross correlation of a two-dimensional data matrix on a data beam using a plurality of spatial light modulators and a spatial accumulator, said method comprising the steps of: (a) buffering a two-dimensional correlation mask data matrix; (b) buffering said two-dimensional data matrix to be correlated; (c) applying the buffered said correlation mask data matrix to the corresponding spatial positions of a two-dimensional spatial light modulator; (d) applying a datum of said two-dimensional data matrix to be correlated, obtained from a predetermined row and column buffer position, to a zero-dimensional spatial light modulator; (e) providing uniform data to all remaining ones of said spatial light modulators; (f) summing into said spatial accumulator the products of the intersecting said correlation mask data matrix and said datum of said two-dimensional data matrix to be correlated, as provided to said two- and said zero-dimensional spatial light modulators, respectively, to thereby produce cumulative two-dimensional resultant data; (g) recirculating the buffered said two-dimensional data matrix to be correlated in a first data matrix direction so as to provide a new datum at said predetermined row and column buffer position; (h) recirculating said cumulative two-dimensional resultant data similarly in said first data matrix direction; (i) repeating steps (c) through (h) for each row or column in said first data matrix direction of said two-dimensional data matrix to be correlated; (j) recirculating the buffered said two-dimensional data matrix to be correlated in a second data matrix direction so as to provide a new datum at said predetermined row and column buffer position; (k) recirculating said cumulative two-dimensional resultant data similarly in said second data matrix direction; and (l) repeating steps (i) through (k) for each row or column in said second data matrix direction of said two-dimensional data matrix to be correlated.

Claim 20

A method of performing a one-dimensional, sliding window cross ambiguity function calculation for one-dimensional data on a data beam using a plurality of spatial light modulators and a spatial accumulator, comprising the steps of: (a) buffering a two-dimensional Fourier transform coefficient mask data matrix; (b) row buffering a one-dimensional reference data template; (c) row buffering an initial portion of a one-dimensional input data stream; (d) applying the respective buffered Fourier transform coefficients of the last matrix column of said Fourier transform coefficient mask data matrix coefficients to respective rows of a first one-dimensional spatial light modulator; (e) applying the respective buffered row of said reference data template data to the corresponding columns of a second one-dimensional spatial light modulator; (f) applying to the respective columns of a third one-dimensional spatial light modulator, corresponding datum from a contiguous portion of said input data stream; (g) providing uniform data to all remaining ones of said spatial light modulators; (h) summing into said accumulator the products of the intersecting said Fourier transform coefficient mask data matrix, said row of said reference data template data, and said contiguous portion of said input data stream as provided to said first, second, and third one-dimensional spatial light modulators, respectively, to thereby produce cumulative two-dimensional resultant data matrix; (i) sequentially shifting the next datum of said input data stream into said buffered portion so as to shift the application of said contiguous portion of said input data stream in a predetermined row data shift direction; (j) recirculating the buffered said Fourier transform coefficient mask data matrix in said predetermined row shift data direction so as to provide the next sequential column of said Fourier transform coefficient mask data matrix to be applied to said first one-dimensional spatial light modulator; (k) shifting said cumulative two-dimensional resultant data matrix in said predetermined row data shift direction and clearing the first matrix column of said resultant data matrix corresponding to the column of said third one-dimensional spatial light modulator to which the next datum of said input data stream is applied; and (l) repeating steps (d) through (k) and sequentially buffering the last matrix column data of said two-dimensional resultant data matrix shifted out during step (k) to form a two-dimensional most recent sliding window cross-ambiguity resultant data matrix.

Claim 21

A method of performing a one-dimensional, sliding window cross ambiguity function calculation for one-dimensional data on a data beam using a plurality of spatial light modulators and a spatial accumulator, comprising the steps of: (a) buffering a two-dimensional Fourier transform coefficient mask data matrix; (b) row buffering a one-dimensional reference data template; (c) row buffering an initial portion of a one-dimensional input data stream; (d) applying the respective buffered Fourier transform coefficients of the last matrix column of said Fourier transform coefficient mask data matrix coefficients to respective rows of a first one-dimensional spatial light modulator; (e) applying the respective buffered row of said reference data template data to the corresponding columns of a second one-dimensional spatial light modulator; (f) applying to the respective columns of a third one-dimensional spatial light modulator, corresponding datum from a contiguous portion of said input data stream; (g) providing uniform data to all remaining ones of said spatial light modulators; (h) summing into said accumulator the products of the intersecting said Fourier transform coefficient mask data matrix, said row of said reference data template data, and said contiguous portion of said input data stream as provided to said first, second, and third one-dimensional spatial light modulators, respectively, to thereby produce cumulative two-dimensional resultant data matrix; (i) sequentially shifting the next datum of said input data stream into said buffered portion so as to shift the application of said contiguous portion of said input data stream in a predetermined row data shift direction; (j) recirculating the buffered said Fourier transform coefficient mask data matrix in said predetermined row shift data direction so as to provide the next sequential column of said Fourier transform coefficient mask data matrix to be applied to said first one-dimensional spatial light modulator; (k) shifting said cumulative two-dimensional resultant data matrix in said predetermined row data shift direction and clearing the first matrix column of said resultant data matrix corresponding to the column of said third one-dimensional spatial light modulator to which the next datum of said input data stream is applied; and (l) repeating steps (d) through (k) and sequentially buffering the last matrix column data of said two-dimensional resultant data matrix shifted out during step (k) to form a two-dimensional most recent sliding window cross-ambiguity resultant data matrix.