Back to results
US 5,515,402 AGranted
Quadrature filter with real conversion
AssigneeHarris Corporation
InventorDavid B. Chester
Life of this patent
Estimating the term…
Abstract
A quadrature decimation filter with real output formed from a mixture of both in-phase and quadrature signals by upconversion and alternating stream combination. The in-phase and quadrature branch computations run simultaneously with a delay of the in-phase data stream relative to the quadrature data stream so that a common filter coefficient may be used simultaneously in nonzero branch computations.
Claims20
A digital filter, comprising:
- (a) a first filter branch with first branch input and first branch output and filter coefficients c.sub.j ;
- (b) a second filter branch with second branch input and second branch output and filter coefficients said c.sub.j ;
- (c) a delay of R input sample intervals at said first branch input where R is a positive integer equal to the ratio of an input sampling rate to an output sampling rate;
- and
- (d) an output formatter with inputs coupled to said first and second branch outputs and alternating passing output samples from said first and second branch outputs at said output sampling rate;
- (e) whereby real data streams x(n) and z(n) entering said delay and said second branch input, respectively, are filtered in said first and second branches by simultaneous multiplications of x(n) and z(n+R) by coefficients c.sub.j and additions or subtractions of the multiplcation products to yield first branch output u(n) and second branch output v(n+R) up converted by one quarter of said output sampling frequency and combined and decimated to form a real filter output u(n), -v(n+R), -u(n+2R), v(n+3R), u(n+4R), -v(n+5R).
3 more elements
A digital filter, comprising:
- (a) a first filter branch with first branch input and first branch output;
- (b) a second filter branch with second branch input and second branch output, and with the same filter coefficients simultaneously used on data streams in said first and second branches;
- (c) a delay of R input sample intervals at said first branch input where R is a positive integer;
- and
- (d) a multiplexer with inputs coupled to said first and second branch outputs and alternating passing output samples from said first and second branch outputs at output sample intervals of length R times the length of said input sample intervals;
- (e) whereby first and second real data streams entering said delay and said second branch input, respectively, are filtered in said first and second branches and up converted at one quarter of an output sampling frequency and combined to form a real filter output.
3 more elements
A method of digitally filtering a complex digital data stream to yield a real digital data stream, comprising the steps of:
- (a) providing first and second real filter branches with said second branch having the same filter coefficients as said first branch;
- (b) applying the real part of a complex digital data stream to an input of said first branch after a delay of R input sampling intervals and thereby forming a first digital data stream with samples every 2R input sampling intervals;
- (c) applying the imaginary part of a complex digital data stream to an input of said second branch and thereby forming a second digital data stream with samples every 2R input sampling intervals and delayed R sampling intervals from said first digital data stream;
- and then
- (d) interleaving said first and second data streams to form a real digital data stream with output sampling interval R times said input sampling interval.
2 more elements
A quadrature decimating FIR filter with complex to real signal conversion comprising:
- plural quadrature filter branches including an in-phase filter branch and a quadrature-phase filter branch,
- at least one of said branches having a delay means, and
- at least two of said branches having common impulse responses;
- and
- conversion means for converting the output of said branches to a real signal.
2 more elements
A method of filtering a complex signal and converting said complex signal into a real signal comprising the steps of:
- (a) providing a complex digital signal having an in-phase component and a quadrature component, each of said components comprised of a series of discrete time samples;
- (b) delaying one of said components by an amount specified by a selected decimation rate;
- (c) sampling each component to thereby produce a sampled in-phase component and a sampled quadrature component;
- (d) scaling said sampled in-phase component and said sampled quadrature component to thereby produce an in-phase filter output and a quadrature filter output;
- (e) upconverting said in-phase filter output and said quadrature filter output to thereby produce a real signal.
2 more elements
A method of filtering a complex signal and converting said complex signal into a real signal comprising the steps of:
- (a) providing a complex digital signal having an in-phase component and a quadrature component, each of said components comprised of a series of discrete time samples;
- (b) providing two shift registers, one dedicated to said in-phase component and one dedicated to said quadrature component, each of said shift registers having an equal number of plural stages, each of said plural stages capable of holding at least one of said discrete time samples;
- (c) providing plural filter coefficients the number of which equals the number of said plural stages with each of said coefficients symmetrically assigned to one of said plural stages in each of said shift registers such that each of said plural stages in each of said registers has a corresponding symmetric partner with the exception of the center plural stage in each of said registers;
- (d) delaying each of said time samples of either the in-phase component or the quadrature component by an amount corresponding to a predetermined decimation rate;
- (e) serially shifting said discrete time samples of said in-phase component and said quadrature component into, through and out of their respective shift registers;
- (f) after every predetermined number of shifts corresponding to a specified decimation rate, sequentially
- (1) adding each plural stage with its symmetric partner,
- (2) multiplying each added output and the center plural stage by their respective coefficients,
- (3) adding each of the multiplication outputs to thereby produce a decimated filtered sample of both the in-phase component and the quadrature component;
- (g) upconverting said decimated filtered samples by successively passing samples of each to thereby produce a real signal;
- and
- (h) complementing every other passed signal of each decimated filter sample .
9 more elements
Dependencies are inferred from the claim text.