Technique for Rapid Determination of Probability of Detection in Pulse Doppler Radars

1. In combination with a pulse doppler radar system having a radar receiver with a plurality of range gates, with a signal processing system containing an analog-to-digital converter, a clutter filter, a fast Fourier transform, an automatic gain control circuit, a doppler filter bank and an envelope detector, said signal processing system being controlled by a system computer, a probability of detection system comprising: a register storing and receiving a reference signal from said fast Fourier transform, said reference signal occuring when said system computer directs an input of a calibrated signal into said radar receiver, said register outputting said reference signal; a first adjusting means receiving and adjusting said reference signal from said register by producing an adjusted reference signal being adjusted for saturation from said analog-to-digital converter, actions by said automatic gain control and notch adjustment from said clutter filter; a first summing means receiving said adjusted reference signal from said first adjusting means, and vectorally adding said adjusting reference signal into said plurality of range gates of a selected doppler filter at the output of said doppler filter bank, said first summing means outputting a summed reference signal to said envelope detector resulting in an envelope detection signal; a first comparison means receiving and comparing said envelope detection signal from said envelope detector, with a detection threshold from a first memory means in said pulse doppler radar system, said first comparison means outputting a detection count for each integration period of said signal processing system; a second memory means receiving said detection count from said first comparison means and outputting a first, second, third and fourth count signals, said first count signal equaling the detection count for the current integration period, the second and third count signals equaling the detection counts received from said first comparison means for the previous two integration periods, and the fourth count signal equaling the total number of range gates tested per integration period; a calculation means receiving said count signals from said second memory means and calculating a probability of detection signal P.sub.D wherein ##EQU4## where N.sub.1, N.sub.2 and N.sub.3 equal respectively said first, second and third count signals, N equals said fourth count signal.

2. A probability of detection system as defined in claim 1 including a second adjusting means receiving and comparing said probability of detection signal from said calculating means with an upper and lower threshold and sending an adjustment input to said first adjusting means to bring said probability of detection signal within said upper and lower threshold.

3. A probability of detection system as defined in claim 2 wherein said first adjusting means comprises: a first programable read-only-memory receiving a signal indicating an amount of attenuation produced by said clutter filter, and producing an adjustment value which, when applied to said reference signal, would compensate for said attenuation; a multiplexer receiving said adjustment value from said first programable read-only-memory and said adjustment input from said second adjusting means, said multiplexer producing an output signal indicating an amount of amplification or attenuation that should be applied to said reference signal; a complex multiplier producing an output signal by receiving and multiplying said reference signal from said register by said output signal from said multiplexer; a compensation unit producing said adjusted reference signal by receiving and adjusting said output signal from said complex multiplier for saturation from said analog-to-digital converter and actions by said automatic gain control.

4. A probability of detection system as defined in claim 3 wherein said calculation means comprises: a Log.sub.2 element receiving said count signals from said second memory means and producing a first, second, third and fourth log signals, said first log signal equaling Log.sub.2 (first count signal), said second log signal equaling Log.sub.2 (second count signal), said third log signal equaling Log.sub.2 (third count signal) and said fourth log signal equaling Log.sub.2 (fourth count signal); and a second summing means producing said probability of detection signal by: adding said first, second and third log signals together to produce a first sum signal, adding said fourth log signal to itself twice to produce a second sum signal, and subtracting said second sum signal from said first sum signal.

5. A probability of detection system as defined in claim 4 wherein said second adjusting means comprises: a second comparison means receiving and comparing said probability of detection signal from said calculation means with an upper and lower threshold, said second comparison means producing an output signal; a second programable read-only-memory receiving said output signal from said second comparison means and producing an output signal indicating a direction and magnitude of deviation of said probability of detection signal from said upper and lower threshold of said second comparison means; and a third programble read-only-memory receiving said output signal from said second programable read-only-memory and producing and sending said adjustment input to said first adjusting means, said adjustment input indicating an amouunt of amplification and attenuation that is required to bring said probability of detection signal within said upper and lower threshold.

6. A technique of indicating a probability of detection by a radar system having a signal processing system, clutter filter, automatic gain control, doppler filter bank and analog-to-digital converter, said technique comprising the steps of: producing a reference signal by injecting a calibrated signal into a radar receiver and measuring an output signal from a fast Fourier transform in said radar's signal processing system; modifying said reference signal for attenuations in said radar's clutter filter, actions by said radar's automatic gain control, and saturations by said radar's analog-to-digital converter, said modifying step resulting in production of a modified reference value; adding said modified reference value to an in-phase amplitude of a selected range gate filter cell in said radar's doppler filter bank, said adding step resulting in a summed reference signal; envelope detecting said summed reference signal along with a stored cell quadrature amplitude from said radar's doppler filter bank, said envelope detecting step resulting in an envelope detection output signal; comparing said envelope detection output signal with a constant false alarm rate value for said selected range gate filter cell and incrementing a detection counter when said envelope detection output exceeds said constant false alarm rate reference value; repeating sid producing, modifying, adding, envelope detecting and comparing steps until all range gate filter cells have been processed in a coherent integration period, said repeating step resulting in a detection count for a current integration period; estimating a probability of detection to provide a probability of detection signal; adjusting said modified reference signal; and repeating said producing, modifying, adding, envelope detecting, comparing and estimating steps when said probability of detection signal is not within a desired range said adjusting step being repeated until a desired detection count is satisfied.

7. A technique as defined in claim 6 wherein said estimating step comprises the steps of: multiplying said detection count from said current integration period from said repeating step by the detection counts obtained in two previous integration periods, said multiplying step producing a first product; obtaining and cubing a number which represents the total number of range gates that are tested each integration period in said radar system, said obtaining and cubing step resulting in a cubed number; and calculating the probability of detection by dividing said first product by said cubed number.