Abstract: Systems and methods of estimating a distance to a cause of a micro-reflection in a CATV network.

Abstract: Systems and methods of estimating a distance to a cause of a micro-reflection in a CATV network.

Abstract: A phase-based approach to estimating the frequency of complex sinusoids in the presence of noise that simultaneously reduces the threshold level and makes it frequency-independent. The approach relies on a filter bank approach: apply the signal to a highly overlapped, four-channel filter bank, detect which channel the signal is in, and apply a phase-based frequency estimation technique to the selected channel output. The computational complexity of the method is kept low by deriving a closed form for the resulting estimator and showing that it leads to an inherent decimation. Further improvement in performance can be obtained by cascading several stages of the four-channel filter bank; the performance tends toward that of the ML method as the number of stages increases.

Abstract: A method for demodulating digital frequency modulation (FM) signals is disclosed. A group of complex-valued discrete-time FM signal samples is initially received. A corresponding first complex product for each of the complex-valued discrete-time FM signal samples is computed, and a corresponding second complex product for each of the first complex values of the complex-valued discrete-time FM signal samples is computed. Subsequently, an inverse tangent of the second complex products are computed to yield an angle for each of the second complex values, wherein each of the angles represents a second-order difference of a phase of the complex-valued discrete-time FM signal samples. Finally, a digital integration is performed to obtain a first-order difference of the phase of the complex-valued discrete-time FM signal samples.

Abstract: Compression of data representing waveform information is dynamically adapted to the waveform by performing a transformation on blocks of samples of the waveform to derive a plurality of coefficients representing a portion of the waveform corresponding to a block of samples and computing quantizer step sizes (quantitative increments of digitization) and quantizer sizes (number of bits for transmission of the coefficients) based on the relative magnitude and dynamic range of the coefficients and a user-specified parameter. The user-specified parameter may be used to set a lower limit on performance such as an upper limit on quantization noise or distortion (in which case data rate is automatically minimized) or a data rate for a communication link (in which case performance relative to quantization noise or distortion is maximized).

Abstract: I/Q imbalance correction is performed directly in the frequency domain by exploiting the symmetry properties of the discrete Fourier transform (DFT) of a real signal, taking the in-phase component as the real part and the quadrature component as the imaginary part. Amplitudes of ordered real and imaginary spectral components are respectively summed with reverse-ordered real and imaginary spectral components, the sums complex multiplied with correction values derived through calibration and selected in accordance with a maximum magnitude spectral component and the results of the complex multiplication summed with the real and imaginary spectral components. Simplified processing architecture and improved performance for a given level of processing power is thus provided.

Abstract: A system is disclosed for estimating the position and velocity of a moving transmitter making use of both Doppler frequency measurements and LBI phase measurements. The frequency measurements and the LBI phase measurements are made from a platform moving through said measurement path. The estimation is based on equation models ##EQU1## These equation models are used in cost evaluations in grid searches to provide initial estimates of parameters defining the position and velocity of the transmitter and the initial estimates are used in nonlinear least squares calculations based on the equation models to provide an estimate of the position and velocity of the transmitter.

Abstract: A multipath signal preprocessor which permits coherent bistatic radar detection with a single omnidirectional antenna is provided. The multipath signal preprocessor separates multipath signals received at an antenna into a strong path signal and a weak path signal. The received multipath signal after filtering, amplifying, and heterodyning is separated into in-phase and quadrature signal components. A constant magnitude signal estimate having approximately the frequency and phase of the strong path signal from the multipath signal is generated. An estimate of the amplitude of the strong path signal is generated from the received multipath signal amplitude. The estimate of the amplitude of the strong path signal is multiplied by the constant magnitude signal estimate having approximately the frequency and phase of the strong path signal to obtain an estimate of the strong path signal.