Digital coherent detector

Claim 2

A digital coherent detector for directly sampling an I.F. signal to obtain its I and Q coefficients comprising: means for sampling and digitizing an analog bandpass signal r(t)=I(t) cos 2πf o t-Q(t) sin 2πf o t at a sampling frequency of f s =2f o /(M-1/2)=2W so that the sin and cos sinosoidal terms alternately go to zero and one, respectively, such that samples r(t m ) are alternately equal to I(t) and then Q(t), where M is an integer, W is the bandwidth of the input signal, I(t) and Q(t) are the in-phase and quadrature coefficient of the signal, and f o is the center frequency of the signal and equals (2M-1)W/2; estimating means connected to the output of said sampling and digitizing means for determining an estimate r(t m ) for the coefficient whose sin or cos sinosoidal term has gone to zero in a given sample from the N preceding and the N succeeding sampled signals; an I output line; a Q output line; and switching means for switching between the I and Q output lines in accordance with the sampling frequency clock pulse alternate samples r(t m ) with appropriate signs, while switching with appropriate sign the estimate signal r(˜t m ) to be paired with each r(t m ) signal to the I or Q output line not connected to take the r(t m ) signal on that clock pulse.

Claim 3

A digital coherent detector as defined in claim 1, wherein said estimating means includes means for determining r(˜t m ) for a given sample by making the determination at the time t m +1/(4f o '), where f o ' is defined identically as f o but has a higher M integer value.

Claim 4

A digital coherent detector as defined in claim 2, wherein said estimating means determines the estimate r(t m +1/4f o ') utilizing the equation ##EQU21## where J(t) is the largest integer less than or equal to (t+1/4W), N is an integer, and s(t-n/2W) is a sampling function.

Claim 5

A digital coherent detector as defined in claim 3, wherein the sampling function s(t-n/2W) for said estimating means is defined as ##EQU22## where q=1-B/W, p=INT (Nq π/e), B is the signal bandwidth, and e is the natural logarithm base.

Claim 6

A digital coherent detector as defined in claims 3 or 4, wherein said estimating means comprises: first 2N+1 stage shift register means for storing sampled signals; a plurality of multipliers, one for each stage in said shift register means for determining the value r(n/2W)s(t-n/2W) for each of the 2N+1 samples from -N+J(t) to +N+J(t); and means for adding the values of r(n/2W)s(t-n/2W) for the 2N+1 samples to yield the estimated value ##EQU23##

Claim 7

A digital coherent detector as defined in claim 5, wherein said first 2N+1 stage shift register means comprises a second and third 2N+1 stage shift registers having reduced bit capacity stages, with each stage from said second shift register paired with a different stage from said third shift register such that the digital word bits representing a given sample r(n/2W) may be divided between the stages forming a given stage pair; and wherein said plurality of multipliers includes a reduced bit capacity multiplier for each stage of said second and third shift registers for determining the value r(n/2W)s(t-n/2W) for that stage; and wherein said adding means includes means for properly adding together the product values from the reduced bit capacity multipliers for a given stage pair which are processing different portions of the same digital word and then progressively adding the different digital words together to obtain ##EQU24##

Claim 8

A digital coherent detector as defined in claim 5, wherein said estimating means includes a first storage means for storing r(t m ) and providing r(t m ) and -r(t m ) on output lines A and B, respectively; and a second storage means for storing the digital value r(t m +1/4f o ') and providing r(t m +1/4f o ') and -r(t m +1/4f o ') on output lines C and D, respectively.

Claim 9

A digital coherent detector as defined in claim 5, wherein said switching means includes means for properly assigning positive and negative signs to r(t m ) and r(t m +1/4f o ') according to the following equations: r(t.sub.m)=(-1).sup.m/2 I(t.sub.m), for m even, r(t.sub.m)=(-1).sup.(m-1)/2 Q(t.sub.m), for m odd, r(t.sub.m +1/4f.sub.o ')=-(-1).sup.m/2 Q(t.sub.m), for m even, and r(t.sub.m +1/4f.sub.o ')=(-1).sup.(m-1)/2 I(t.sub.m), for m odd.

Claim 10

A digital coherent detector as defined in claim 7, wherein said switching means comprises first and second four-input commutating switches with their inputs connected to the A, B, C, and D lines from said first and second storage means, and with their outputs connected to said I and Q output lines, respectively, for applying thereto r(t m ), -r(t m ), r(t m +1/4f o ') and -r(t m +1/4f o '), as appropriate.

Claim 11

A digital coherent detector as defined in claim 9, wherein the lines A, B, C, and D are connected to the inputs for said first commutating switch in the order A, C, B, D, and are connected to the inputs for said second commutating switch in the order D, A, C, B.

Claim 12

A digital coherent detector for directly sampling an I.F. signal to obtain its I and Q coefficients comprising: means for sampling and digitizing an analog bandpass signal r(t)=I(t) cos 2πf o t-Q(t) sin 2πf o t at a sampling frequency of f s =2f o /(M-1/2)=2W so that the sin and cos sinosoidal terms alternately go to zero and one, respectively, such that samples r(t m ) are alternately equal to I(t) and then Q(t), where M is an integer, W is the bandwidth of the input signal, I(t) and Q(t) are the in-phase and quadrature coefficients of the signal, and f o is the center frequency of the signal and equals (2M-1)W/2; a finite impulse response digital filter connected to the output of said sampling and digitizing means for determining an estimate r(˜t m ) for the coefficient whose sinosoidal term has gone to zero for a given sample r(t m ), utilizing the equation ##EQU25## where J(t) is the largest integer less than or equal to t+1/4W, N is an integer number of samples, and s(t-n/2W) is a sampling function, and wherein the times used for the determination are t=t m +1/4f o ' where f o '=(2M'-1)W/2, with M' being an integer >>M; an I output line; a Q output line; and switching means for properly setting the signs for r(t m ) and r(˜t m ) and alternately switching these terms between said I and Q output lines in accordance with a clock pulse.

Claim 13

A digital coherent detector for directly sampling an I.F. signal to obtain its I and Q coefficients comprising: means for sampling and digitizing an analog bandpass signal r(t)=I(t) cos 2πf o t-Q(t) sin 2πf o t at a sampling frequency of f s =2f o /(M-1/2)=2W so that the sin and cos sinosoidal terms alternately go to zero and one, respectively, such that samples r(t m ) are alternately equal to I(t) and then Q(t), where M is an integer, W is the bandwidth of the input signal, I(t) and Q(t) are the in-phase and quadrature coefficients of the signal, and f o is the center frequency of the signal and equals (2M-1)W/2; a finite impulse response digital filter connected to the output of said sampling and digitizing means for determining an estimate r(˜t m ) for the coefficient whose sinosoidal term has gone to zero for a given sample r(t m ), utilizing the equation ##EQU25## where J(t) is the largest integer less than or equal to t+1/4W, N is an integer number of samples, and s(t-n/2W) is a sampling function, and wherein the times used for the determination are t=t m +1/4f o ' where f o '=(2M'-1)W/2, with M' being an integer >>M; an I output line; a Q output line; and switching means for properly setting the signs for r(t m ) and r(˜t m ) and alternately switching these terms between said I and Q output lines in accordance with a clock pulse.

Claim 14

A digital coherent detector as defined in claim 11, wherein said FIR digital filter comprises: 2N+1 stage shift register means for storing and shifting sampled signals; a plurality of ROM multipliers, one for each stage in said shift register means for determining the value r(n/2W)s(t-n/2W) for each of the samples from -N+J(t) to +N+J(t); means for adding the values of r(n/2W)s(t-n/2W) for the samples from -N+J(t) to +N+J(t) to yield r(t m +1/4f o '); a first buffer register connected to the appropriate stage in said 2N+1 stage shift register means for storing r(t m ) and providing +r(t m ) and -r(t m ) outputs to said switching means; a second buffer register for storing r(t m +1/4f o ') and providing r(t m +1/4f o ') and -r(t m +1/4f o ') outputs to said switching means; and wherein said switching means applies the appropriate outputs from r(t m ), -r(t m ), r(t m +1/4f o '), and -r(t m +1/4f o ') to said I and Q output lines.