Apparatus for and method of, analysing signals

Claim 2

A method as claimed in claim 1, which includes, as a further step between steps (iii) and (iv), the step of: selecting the dominant spectral components of the first frequency components or the second frequency components, corresponding to the dominant constituents of the original signal, the subsequent steps being effected on the dominant spectral components.

Claim 3

A method as claimed in claim 1, which includes the additional step: (vii) obtaining the initial phase of each constituent of the original signal from the respective frequency and argument of the respective frequency components.

Claim 4

A method as claimed in claim 3, which includes the additional step: (viii) obtaining the amplitude of at least the first frequency component or the second frequency component, and obtaining the amplitude of each respective constituent of the original signal from the respective frequency, initial phase and amplitude of the respective frequency component.

Claim 5

A method as claimed in claim 2, wherein said further step comprises comparing the magnitude of each component with the magnitude of subsequent and preceding components, to determine whether that component is the greatest, and comparing that component with a predetermined threshold value, to determine whether that component is greater than the threshold value.

Claim 6

A method as claimed in claim 2, wherein said further step comprises obtaining the difference between the argument of one component and the argument of the preceding component or the subsequent component, taking the difference between those two arguments, and determining whether that difference is equal to the value (1-1/N).pi., within predetermined limits, to determine the presence of a dominant spectral component.

Claim 7

A method as claimed in claim 6, where said further step additionally comprises comparing the amplitude of each component with the amplitude of the subsequent and preceding components, to determine whether that component is the greatest and comparing the magnitude of that component with a predetermined threshold value to determine whether that component is greater than the threshold 36 value, and to determine that a dominant spectral component is present only when the magnitude of the component is greater than the magnitude of subsequent and preceding components and greater than the threshold value, and when the argument between that component and a subsequent or preceding component is equal to said value within a tolerance.

Claim 8

A method as claimed in claim 1, wherein the step (vi) comprises for each frequency component: making the respective phase difference modular and positive with respect to on; and dividing the phase difference by 2.pi. and by the shift applied to the shifted signal, to give the frequency of the respective component.

Claim 9

A method as claimed in claim 3, wherein step (vii) comprises for each dominant spectral component; taking the argument thereof and adding the value .pi./2 to obtain a first value; multiplying the respective frequency by the duration of a window applied to the original signal, subtracting the corresponding index, and multiplying the resultant by a quantity (1-1/N).pi. to obtain a second value; subtracting the second value from the first value, and making the resultant modular with respect to 2.pi. to give the initial phase of the corresponding frequency component.

Claim 10

A method as claimed in claim 4, wherein step (viii) comprises, for each frequency component: obtaining the product of the corresponding frequency and the duration of a window applied to the original signal; subtracting the respective index from that product to obtain a first value, and adding the respective index 37 to that product to obtain a second value; obtaining the magnitude and argument of three elements making up the respective dominant spectral component from the first value, with the second element being calculated from the first value plus the increment 1 and with the third element being calculated from the first value minus the increment 1; obtaining the magnitude and argument of fourth, fifth and sixth elements of the respective dominant component from the second value, with the fifth element being calculated from the second value plus the increment 1, and with the sixth element being calculated from the second value minus the increment 1; summing the said six elements and obtaining a resultant magnitude and argument; and dividing the amplitude of the respective dominant spectral component by said resultant magnitude, to obtain the amplitude of the respective constituent of the original signal.

Claim 11

A method as claimed in claim 1, 3 or 4, wherein the shift applied to the signal is a time shift.

Claim 12

A method as claimed in claim 2, wherein the shift applied to the signal is a time shift.

Claim 14

An apparatus as claimed in claim 13, which includes: (vii) a dominant spectral components detector, which is connected to an output of one of said first and second transform means, to determine dominant spectral components of the signal.

Claim 15

An apparatus as claimed in claim 14, which includes: (viii) initial phase obtaining means connected to the argument obtaining means and to the frequency obtaining means, for obtaining the initial phases of the dominant constituents.

Claim 16

An apparatus as claimed in claim 15, which includes: (ix) a first amplitude obtaining means, connected to one of the first and second transform means, for obtaining the amplitudes of the dominant spectral components; and (x) a second amplitude obtaining means, connected to the first amplitude obtaining means, the initial phase obtaining means and the frequency obtaining means, for obtaining the amplitudes of each said constituent from the respective frequency, the respective initial phase and the amplitude of the respective dominant spectral component.

Claim 17

An apparatus as claimed in claim 16, wherein the argument obtaining means comprises a rectangular-to-polar converter connected to an output of the first transform means, for obtaining the magnitude and argument of the first frequency components, and an argument obtaining unit, connected to an output of the second transform means and to the phase difference obtaining means, the rectangular-to-polar converter also being connected to the phase difference obtaining means and incorporating the first amplitude obtaining means.

Claim 18

An apparatus as claimed in claim 17, wherein the rectangular-to-polar converter is connected via a first selector to the phase difference obtaining means, ens wherein the second transform means is connected via a second selector to the phase difference obtaining means.

Claim 19

An apparatus as claimed in claim 18, wherein the first and second transform means have their outputs connected to first and second output buffer registers, the first transform means being connected to the first output buffer register via the rectangular-to-polar converter.

Claim 20

An apparatus as claimed in claim 17, 18 or 19, wherein the rectangular-to-polar converter derives the square of the amplitude of the first frequency components, and wherein a square root unit is provided between the rectangular-to-polar converter and the second amplitude obtaining means.

Claim 21

An apparatus as claimed in claim 16, which includes a window coefficient generator, for applying a predetermined window to both an original signal and a shifted signal.

Claim 22

An apparatus as claimed in claim 21, which further includes a window coefficient register, to which the window coefficient generator is connected, and first and second window multipliers, which are connected respectively to the inputs of the first and second transform means, and are connected to the window coefficient register.

Claim 23

An apparatus as claimed in claim 22, which includes first and second input buffer registers, connected respectively to inputs of the first and second window multipliers.

Claim 24

An apparatus as claimed in claim 21, 22 or 23, which includes an antialiasing filter and an analogue-to-digital converter, which are connected in series, and are connected to an input of the first transform means, and via the delay means to the input of the second transform means.

Claim 25

An apparatus as claimed in claim 14, wherein the dominant spectral components detector comprises: a first comparator having two inputs for the magnitude of one frequency component and the magnitude of a preceding frequency component; a second comparator having two inputs for the magnitude of said one frequency component and the magnitude of a subsequent frequency component; a third comparator having two inputs for the magnitude of said one frequency component and a threshold value; an AND gate having three inputs connected to outputs of three comparators; an active transition detector connected to an output of the AND gate, whereby the active transition detector delivers a signal if the magnitude of said one frequency component is greater than the magnitude of the preceding and subsequent frequency components and 41 greater than the threshold value.

Claim 26

An apparatus as claimed in claim 14, wherein the dominant spectral components detector comprises: a subtractor having an input for an argument of one frequency component, and an input for the argument of an adjacent frequency component; and a comparison means connected to an output of the subtractor for comparing the output of the subtractor with the quantity (1-1/N).pi., the comparison means sending a signal indicative of a dominant spectral component if the difference between the output of the subtractor and that value is less than a predetermined amount.

Claim 27

An apparatus as claimed in claim 26, wherein the comparison means comprises one comparator for comparing the output of the subtractor with an upper boundary value, and a second comparator for comparing the output of the subtractor with a lower boundary value.

Claim 28

An apparatus as claimed in claim 26 or 27, wherein the subtractor is connected to the comparison means via a circuit for adding the quantity 2.pi. to the output of the subtractor, if the subtractor output is negative, so that the comparison means receives a positive value.

Claim 29

An apparatus means as claimed in claim 26 or 27 which further comprises: a first comparator having some input for the magnitude of a frequency component and another input for the magnitude of a preceding frequency component; a second comparator having an input for the magnitude of a frequency component and the magnitude of a subsequent frequency component; a third comparator having an input for the magnitude of a frequency component and an 42 input for a threshold value; an AND gate having three inputs connected to outputs of the first, second and third comparators; an active transition detector, connected to an output of the AND gate; and an output AND gate having an input connected to the output of the active transition detector and an input connected to an output of the comparison means; whereby, in use, said output AND gate provides a signal indicative of a dominant spectral component, when the magnitude of a frequency component is greater than or equal to the magnitude of subsequent preceding components and is greater than a predetermined threshold value, and when the difference between the argument of that frequency component and the argument of a subsequent or preceding component is equal to the value (1-1/N).pi. within a predetermined tolerance.

Claim 30

An apparatus as claimed in claim 25, 26 or 27, which additional includes an increment counter, for recording the number of dominant spectral components, and a recorder, for recording the index values corresponding to the dominant spectral components.

Claim 31

An apparatus as claimed in claim 13, wherein the frequency obtaining means comprises: a circuit for adding the value 2.pi. to the respective phase difference when that phase difference is less than 0, to produce an adjusted phase difference value; and a divider connected to the output of said circuit for dividing the adjusted phase difference value by 2.pi. and the duration of the timing delay.

Claim 32

An apparatus as claimed in claim 15, wherein the initial phase obtaining means comprises: first, second and third input for respectively, the argument of each dominant spectral 43 component, the index corresponding to that dominant spectral component and the frequency corresponding to that dominant spectral component; a first adder having an input connected to said first input, and another input supplied with the value .pi./2; a first multiplier unit connected to the third input and to a fixed value NT; a first subtractor having a negative input connected to said second input and a positive input connected to an output of the first multiplier unit; a second multiplier unit having an input connected to the output of the subtractor and an input for the value (1-1/N).pi.; a second subtractor having a positive input connected to an output of the adder and a negative input connected to the output of the second multiplier unit; and a modular 2.pi. unit connected to an output of the second subtractor, the output of the modular 2.pi. unit being the initial phase for the corresponding constituent of the original signal.

Claim 33

An apparatus as claimed in claim 16, wherein the second amplitude obtaining means comprises: a first input for the amplitude of the respective frequency component; a second input for the product of the respective frequency and the duration of a window for the signal; a third input for the index value, a first subtractor having a positive input connected to said second input and a negative input connected to said third input; a first adder having positive inputs connected to said second and third inputs; first and second increment units connected respectively to the outputs of the subtractor and the first adder, for adding the quantity 1; first and second decrement units connected respectively to outputs of the subtractor and the first adder for subtracting the quantity 1; six pairs of magnitude and argument 44 deriving networks, the first pair of which is convected to the output of the subtracter, the second pair of which is connected to the output of the first increment unit, the third pair of which is connected to the output of the first decrement unit, the fourth pair of which is connected to the output of the adder, the fifth pair of which is connected to the output of the second increment unit, the sixth pair of which is connected to the output of the second decrement unit, and each pair of which is connected to a respective multiplication unit; a summation unit having inputs connected to the outputs of said multiplication units and an output; a square, sum and square root circuit connected to the output of the summation unit; and a divider connected to the output of said square, sum and square root circuit and to the first input, for dividing the amplitude of the respective frequency component by the output of said circuit to derive the amplitude of the respective constituent of the original signal.

Claim 34

An apparatus as claimed in claim 33, wherein each of the magnitude derivation network includes means for determining the quantity .gamma.Sin(.pi.S)/Sin(.pi.S/N), where S is the input to the network, N is the number of samples of the original signal and .gamma. is a constant dependent on the type of window applied to the original signal, except when the input signal S is equal to zero, when the network provides the output N.gamma..

Claim 35

An apparatus as claimed in claim 33, wherein each of the argument derivation networks comprises: a multiplication unit having an input provided with the input signal for the network, and another input provided with the value (1-1/N).pi.; a respective adder having a first input connected to the output of the multiplication unit, and a second input supplied with the sum of the initial phase of the shifted signal and the respective phase difference less than the quantity .pi./2; and a function unit connected to an output of the adder, the function unit in the first, second and third argument derivation networks calculating the function EXP(jP), P being the output of the adder, and for the fourth, fifth and sixth argument calculation networks calculates the function EXP(-jQ), Q being the output of the adder, the output of the function unit being the output of the respective argument calculation network.

Claim 36

An apparatus as claimed in claim 13, 15 or 16, wherein the delay means applies a time delay to the signal. 46