Filter that minimizes in-band noise and maximizes detection sensitivity of exponentially-modulated signals

A system that demodulates an exponentially-modulated signal by minimizing noise in a signal band of the exponentially-modulated signal while maximizing a detection sensitivity resulting in a negative noise figure defined by a ratio of input signal-to-noise ratio (SNRi) to an output signal-to-noise ratio (SNRo) that is less than one, said system compressing the noise in the signal band, said system comprising:

• a first linear time-invariant (LTI) filter comprising:
• a first pair of parallel paths having a common input into which is fed the exponentially modulated signal having the input signal-to-noise ratio (SNRi) and wherein at least one of said first parallel paths comprises a time delay for delaying the exponentially-modulated signal passing therethrough, each of said first parallel paths comprising a respective output signal and wherein said respective output signals are delayed in time from each other by a predetermined parameter, said parallel paths and said time delay comprising only linear components, said linear components minimizing a threshold that is normally present in exponentially-modulated signal demodulators that use non-linear components;
• a first summing network for receiving said respective output signals from said first pair of parallel paths and summing said respective output signals to generate an output that is a function of a rate change of a carrier frequency of said exponentially-modulated signal, said output comprising impulses for abrupt changes in the carrier frequency corresponding to data transitions that form the baseband of said exponentially-modulated signal;
• and
• wherein said first LTI filter is operative on any modulation format of said exponentially modulated signal and generates said impulses while compressing noise in the signal band;
• and
• a second LTI filter comprising:
• a second pair of parallel paths having a common input into which said impulses are fed and wherein at least one of said second parallel paths comprises a time delay for delaying the impulses passing therethrough, each of said second parallel paths comprising a respective output signal and wherein said respective output signals of said second pair of parallel paths are delayed in time from each other by said predetermined parameter, said second pair of parallel paths and said time delay comprising only linear components, said linear components minimizing the threshold that is normally present in exponentially-modulated signal demodulators that use non-linear components;
• and
• a second summing network for receiving said respective output signals from said second pair of parallel paths and summing said respective output signals to further increase amplitudes of said impulses for abrupt changes in the carrier frequency corresponding to data transitions that form the baseband of said exponentially-modulated signal while further compressing noise in the signal band to form a second LTI filter output and wherein said second LTI filter output comprises said output signal-to-noise ratio (SNRo) that is greater than the input signal-to-noise ratio (SNRi).

A system that demodulates an exponentially-modulated signal by minimizing noise in a signal band of the exponentially-modulated signal while maximizing a detection sensitivity resulting in a negative noise figure defined by a ratio of input signal-to-noise ratio (SNRi) to an output signal-to-noise ratio (SNRo) that is less than one, said system compressing the noise in the signal band, said system comprising:

• a filter comprising:
• a first frequency selective network and a second frequency selective network using only linear components and which form a frequency domain derivative operator that generates an output that is a function of a rate of change of a carrier frequency of the exponentially modulated signal at an input to said filter, the exponentially modulated signal having the input signal-to-noise ratio (SNRi) and wherein said linear components minimizing a threshold that is normally present in exponentially modulated signal demodulators;
• said first frequency selective network tuned above a center frequency of said filter and said second frequency selective network tuned below the center frequency of said filter, said first frequency selective network and said second frequency selective network being configured in parallel and both receiving the exponentially modulated signal as a first input signal, said first frequency selective network generating a first output signal and said second frequency selective network generating a second output signal, and wherein said first output signal and said second output signal are 180 degrees out of phase;
• and
• a summing network for receiving said first and second output signals and summing said first and second output signals to generate a filter output signal comprising impulses for abrupt changes in the carrier frequency corresponding to data transitions that form the baseband of said exponentially modulated signal and wherein said filter output signal comprises said output signal-to-noise ratio (SNRo) that is greater than the input signal-to-noise ratio (SNRi);
• and
• wherein said filter is linear time-invariant (LTI) and is operative on any modulation format of said exponentially modulated signal and generates said impulses while compressing noise in the signal band.

A system that demodulates an exponentially-modulated signal by minimizing noise in a signal band of the exponentially-modulated signal while maximizing a detection sensitivity resulting in a negative noise figure defined by a ratio of input signal-to-noise ratio (SNRi) to an output signal-to-noise ratio (SNRo) that is less than one, said system compressing noise in the signal band, said system comprising:

• a first linear time-invariant (LTI) filter comprising:
• a first frequency selective network and a second frequency selective network using only linear components and which form a frequency domain derivative operator that generates an output that is a function of a rate of change of a carrier frequency of the exponentially modulated signal at an input to said first filter, the exponentially modulated signal having the input signal-to-noise ratio (SNRi) said linear components minimizing a threshold that is normally present in exponentially modulated signal demodulators;
• said first frequency selective network tuned above a center frequency of said first filter and said second frequency selective network tuned below the center frequency of said first filter, said first frequency selective network and said second frequency selective network being configured in parallel and both receiving the exponentially modulated signal as a first input signal, said first frequency selective network generating a first output signal and said second frequency selective network generating a second output signal, and wherein said first output signal and said second output signal are 180 degrees out of phase;
• and
• a first summing network for receiving said first and second output signals and summing said first and second output signals to generate impulses for abrupt changes in the carrier frequency corresponding to data transitions that form the baseband of said exponentially modulated signal;
• and
• wherein said filter is operative on any modulation format of said exponentially modulated signal and generates said impulses while compressing noise in the signal band;
• and
• a second LTI filter comprising:
• a third frequency selective network and a fourth frequency selective network using only linear components and which also form a frequency domain derivative operator that receive said impulses from said first filter and obtains a second derivative of said impulses, said second filter further compressing noise in the signal band and further increasing amplitudes of said impulses;
• and
• said third frequency selective network tuned above the center frequency of said first filter and said fourth frequency selective network tuned below the center frequency of said first filter, said third frequency selective network and said fourth frequency selective network also being configured in parallel and each having outputs that are also 180 degrees out of phase and which are fed to a second summing network for generating a second derivative of said impulses which comprises a second LTI filter output and wherein said second LTI filter output comprises said output signal-to-noise ratio (SNRo) that is greater than the input signal-to-noise ratio (SNRi).

A method for demodulating an exponentially-modulated signal by minimizing noise in a signal band of the exponentially-modulated signal while maximizing a detection sensitivity for generating a negative noise figure defined by a ratio of an input signal-to-noise ratio (SNRi) to an output signal-to-noise ratio (SNRo) that is less than one, said method for compressing the noise in its signal band, said method comprising:

• feeding the exponentially-modulated signal, having the input signal-to-noise ratio (SNRi), into a linear time-invariant (LTI) filter comprising a pair of parallel paths having a common input and wherein at least one of said parallel paths comprises a time delay for delaying the exponentially-modulated signal passing therethrough, each of said parallel paths comprising a respective output signal and wherein said respective output signals are delayed in time from each other by a predetermined parameter, said parallel paths and said time delay comprising only linear components, said linear components minimizing a threshold that is normally present in exponentially-modulated signal demodulators that use non-linear components, said filter being operative on any modulation format of the exponentially-modulated signal;
• summing said respective output signals to generate an LTI filter output that is a function of a rate change of a carrier frequency of the exponentially-modulated signal, said output comprising impulses for abrupt changes in the carrier frequency corresponding to data transitions that form the baseband of said exponentially-modulated signal;
• and
• compressing noise in said signal band and wherein said LTI filter output comprises said output signal-to-noise ratio (SNRo) that is greater than the input signal-to-noise ratio (SNRi).

A method for demodulating an exponentially-modulated signal by minimizing noise in a signal band of the exponentially-modulated signal while maximizing a detection sensitivity for generating a negative noise figure defined by a ratio of an input signal-to-noise ratio (SNRi) to an output signal-to-noise ratio (SNRo) that is less than one, said method for compressing the noise in its signal band, said method comprising:

• feeding the exponentially-modulated signal, having the input signal-to-noise ratio (SNRi), into a first linear time-invariant (LTI) filter comprising a pair of parallel paths having a common input and wherein at least one of said parallel paths comprises a time delay for delaying the exponentially-modulated signal passing therethrough, each of said parallel paths comprising a respective output signal and wherein said respective output signals are delayed in time from each other by a predetermined parameter, said parallel paths and said time delay comprising only linear components, said linear components minimizing a threshold that is normally present in exponentially-modulated signal demodulators that use non-linear components, said filter being operative on any modulation format of the exponentially-modulated signal;
• summing said respective output signals to generate a first LTI filter output that is a function of a rate change of a carrier frequency of the exponentially-modulated signal, said output comprising impulses for abrupt changes in the carrier frequency corresponding to data transitions that form the baseband of said exponentially-modulated signal;
• and
• compressing noise in said signal band;
• feeding said impulses of said first LTI filter output into a second LTI filter comprising a second pair of parallel paths having a common input and wherein at least one of said second parallel paths comprises a second time delay for delaying the impulses passing therethrough, each of said second parallel paths comprising a respective output signal and wherein said respective output signals of said second parallel paths are delayed in time from each other by said predetermined parameter, said second parallel paths and said second time delay comprising only linear components, said linear components minimizing said threshold that is normally present in exponentially-modulated signal demodulators that use non-linear components;
• and
• summing said respective output signals of said second parallel paths to generate further impulses while compressing noise in the signal band and further increasing amplitudes of said impulses to form a second LTI filter output, and wherein said second LTI filter output comprises said output signal-to-noise ratio (SNRo) that is greater than the input signal-to-noise ratio (SNRi).

A method for demodulating an exponentially-modulated signal by minimizing noise in a signal band of the exponentially-modulated signal while maximizing a detection sensitivity for generating a negative noise figure defined by a ratio of an input signal-to-noise ratio (SNRi) to an output signal-to-noise ratio (SNRo) that is less than one, said method for compressing noise in its signal band, said method comprising:

• feeding the exponentially modulated signal, having the input signal-to-noise ratio (SNRi) into a linear time-invariant (LTI) filter comprising first and second frequency selective networks that use only linear components for minimizing a threshold that is normally present in exponentially modulated signal demodulators, said first frequency selective network being tuned above a center frequency of said filter and said second frequency selective network being tuned below the center frequency of said filter, said first frequency selective network and said second frequency selective network being configured in parallel and both receiving the exponentially modulated signal as a first input signal, said first frequency selective network generating a first output signal and said second frequency selective network generating a second output signal, and wherein said first output signal and said second output signal are 180 degrees out of phase, said filter being operative on any modulation format of the exponentially-modulated signal;
• summing said first and second output signals to generate an LTI filter output comprises impulses for abrupt changes in the carrier frequency corresponding to data transitions that form the baseband of said exponentially modulated signal, said impulses corresponding to a frequency domain derivative of said exponentially modulated signal which is a function of a rate of change of the carrier frequency of the exponentially modulated signal;
• and
• compressing noise in said signal band and wherein said LTI filter output comprises said output signal-to-noise ratio (SNRo) that is greater than the input signal-to-noise ratio (SNR).