Filter and method for whitening digitally generated noise

Claim 1

A method of generating a whitened noise signal, the method comprising the application of an algorithm to a signal processing methodology: I. wherein the signal processing methodology comprises: (A) using a signal source to generate a sequence of digital electrical signals corresponding to a sequence of uniformly distributed independent random numbers, the uniform distribution and independence of said numbers corresponding to a flat spectral response; (B) using a signal processor coupled to the signal source to process said sequence of digital electrical signals according to an algorithm, detailed below, to yield a sequence of prewhitened digital electrical signals; and (C) applying the prewhitened digital electrical signals, seriatim, to a digital-to-analog converter that is coupled to the signal processor to produce a time varying analog electrical signal therefrom, the digital-to-analog converter performing the following functions: (a) converting each of said prewhitened digital electrical signals into an analog signal having an amplitude corresponding thereto; (b) holding said analog signal at said corresponding amplitude for a finite time interval; and (c) introducing a ripple in a spectral distribution of the time varying analog electrical signal; and II. wherein the algorithm comprises repeating the following two steps with different of the digital electrical signals as first and second data: (i) weighting the first and second data in a one to eight ratio; and (ii) computing a difference between the weighted first and second data and producing one of said prewhitened digital electrical signals therefrom.

Claim 2

A method of generating a whitened noise signal, the method comprising the application of an algorithm to a signal processing methodology: I. wherein the signal processing methodology comprises: (A) using a signal source to generate a sequence of digital electrical signals corresponding to a sequence of uniformly distributed independent random numbers, the uniform distribution and independence of said numbers corresponding to a flat spectral response; (B) using a signal processor coupled to the signal source to process said sequence of digital electrical signals according to an algorithm, detailed below, to yield a sequence of prewhitened digital electrical signals; and (C) applying the prewhitened digital electrical signals, seriatim, to a digital-to-analog converter that is coupled to the signal processor to produce a time varying analog electrical signal therefrom, the digital-to-analog converter performing the following functions: (a) converting each of said prewhitened digital electrical signals into an analog signal having an amplitude corresponding thereto; (b) holding said analog signal at said corresponding amplitude for a finite time interval; and (c) introducing a ripple in a spectral distribution of the time varying analog electrical signal; and II. wherein the algorithm comprises repeating the following two steps with different of the digital electrical signals as first and second data: (i) weighting the first and second data in a one to eight ratio; and (ii) computing a difference between the weighted first and second data and producing one of said prewhitened digital electrical signals therefrom.

Claim 4

An apparatus comprising: a source for providing sampled digital data signals, the sampled data signals corresponding to uniformly distributed independent N-bit digital noise words, the uniform distribution and independence of said noise words corresponding to a flat spectral response, but the sampled nature thereof introducing a sin(x)/x ripple thereto; a bit shifter having an input for receiving a sampled digital data signal from the source and having an output, the output providing an output digital data signal corresponding to the input data signal shifted by three bits, the bit shifter thereby weighting the value of the input data signal by a factor of eight; and a processing unit having a first input coupled to the output of the bit shifter and having a second input coupled to the source, said processing unit having an output to which is provided a processed digital data signal corresponding to a difference between the digital data signals applied to the two processing unit inputs from the bit shifter and from the source.

Claim 5

An apparatus comprising: a source for providing sampled digital data signals, the sampled data signals corresponding to uniformly distributed independent N-bit digital noise words, the uniform distribution and independence of said noise words corresponding to a flat spectral response, but the sampled nature thereof introducing a sin(x)/x ripple thereto; a bit shifter having an input for receiving a sampled digital data signal from the source and having an output, the output providing an output digital data signal corresponding to the input data signal shifted by three bits, the bit shifter thereby weighting the value of the input data signal by a factor of eight; and a processing unit having a first input coupled to the output of the bit shifter and having a second input coupled to the source, said processing unit having an output to which is provided a processed digital data signal corresponding to a difference between the digital data signals applied to the two processing unit inputs from the bit shifter and from the source.