Abstract: Embodiments of the present disclosure provide mechanisms that enable designing an FIR filter that would have a guaranteed globally optimal magnitude response in terms of the minimax optimality criterion given a desired weight on the error in the stopband versus the passband. Design of such a filter is based on a theorem (“characterization theorem”) that provides an approach for characterizing the global minimax optimality of a given FIR filter h[n], n=0, 1, . . . , N, where optimality is evaluated with respect to a magnitude response of this filter, |H(ej?)|, as compared to the desired filter response, D(?), which is unity in the passband and zero in the stopband. The characterization theorem enables characterizing optimality for both real-valued and complex-valued filter coefficients, and does not require any symmetry in the coefficients, thus being applicable to all non-linear phase FIR filters.

Abstract: An apparatus for determining the presence of a tone in an input signal includes memory circuitry and data processing circuitry coupled to the memory circuitry. The data processing circuitry is operative to receive multiple samples of the input signal, and to determine a first value at least in part by multiplying each of the samples by respective ones of a first set of values for an impulse response and summing the results. The data processing system is also operative to determine a second value at least in part by multiplying each of a portion of the samples by respective ones of a second set of values for the impulse response and summing the results. The data processing system is operative to determine the power of the tone in the multiple samples of the input signal at least in part by utilizing the first value and the second value.

Abstract: An apparatus for determining the presence of a tone in an input signal includes memory circuitry and data processing circuitry coupled to the memory circuitry. The data processing circuitry is operative to receive multiple samples of the input signal, and to determine a first value at least in part by multiplying each of the samples by respective ones of a first set of values for an impulse response and summing the results. The data processing system is also operative to determine a second value at least in part by multiplying each of a portion of the samples by respective ones of a second set of values for the impulse response and summing the results. The data processing system is operative to determine the power of the tone in the multiple samples of the input signal at least in part by utilizing the first value and the second value.

Abstract: Methods and apparatus to decode dual-tone signals are disclosed. An example receiver to decode a dual-tone signal includes a tone detector to detect a start of a first tone, a first counter to count first cycles of the first tone, a second counter to count second cycles of a system clock while the first counter is counting and the first count is less than a first threshold, state control logic to start the second counter counting third cycles of the clock when a time period elapses, the third count being substantially equal to the second count, the first counter to count fourth cycles of a second tone while the third cycles are counted, and a decoder to compare the fourth count to a second threshold to identify an event represented by the signal.

Abstract: A tone detector is disclosed that is realizable in digital embodiment on a single integrated circuit die and does not require external components, such as a discrete capacitor. An input connects to a comparator, which in turn connects to one or more edge detectors and a flip flop. The edge detector outputs a pulse responsive to a detected edge. A counter is reset by the pulses from the edge detectors thereby preventing the counter from reaching a maximum value, which would otherwise be output from the counter and provided to a flip flop to clock in the comparator output at the D input to the flip flop. In operation, the comparator generates a rail to rail signal responsive to a received tone, which in turn is clocked through the flip flop as a logic high output indicating presence of a tone.

Abstract: Systems and methods are disclosed for computing a power spectral density of a signal. One such method computes a Fast Fourier Transform (FFT) of a first signal segment and an FFT of a second signal segment. The first and second segments are adjacent and non-overlapping in time, and the second segment follows the first. The first FFT is multiplied by a complex conjugate of the second FFT. A first complex value is computed from this first product. An FFT of a third signal segment is computed. The second and third segments are adjacent and non-overlapping in time. The third segment follows the second segment. The second FFT is multiplied by a complex conjugate of the third FFT. A second complex value is computed from this second product. The first and second complex values are added, and the power spectral density of the sum is computed.

Abstract: A method for operating a data storage medium when changing from an operating mode to a directly subsequent power-saving quiescent mode, where the operating mode effects a transmission delay for the last item of information which is to be transmitted, so that immediately after the last item of information which is to be transmitted has been transmitted the quiescent mode is activated and the maximum permissible power consumption is observed.

Abstract: An apparatus and a method for detecting the energy of a tone signal. A pre-process unit takes absolute values of transmitted samples and outputs the values to a delay unit for delaying the transmitted samples according to a predetermined delay interval. An operation unit subtracts the value calculated by multiplying the magnitudes of the samples transmitted before and after the burst detection from the square value of the magnitude of the sample at the time of burst detection among the delayed transmitted samples. A detection unit detects the burst by confirming whether peaks corresponding to the number of the samples transmitted before the sample of the burst detection are detected by a signal of an value input from the operation unit. Therefore, the present invention can reduce a probability of a false alarm caused by an impulse error and burst errors and thus perform stable burst detection.

Abstract: This invention relates to detecting a pilot tone in an optical fiber. In the method described in the invention, computational complexity is reduced significantly by almost completely removing the need to carry out multipli-cations. To calculate certain spectral components of DFT (Discrete Fourier Transform) only a few multiplications are required. The idea is that for de-tecting one pilot tone, it is adequate to calculate one spectral component of DFT if a specific ratio can be chosen between the pilot tone frequency and the sample rate of the receiver.

Abstract: A signal detector for detecting a signal having a first nominal frequency, F1 and a second nominal frequency F2, the signal intermixed with an interfering signal to form a composite signal, the signal detector including a first and a second band pass filter centered at F1 and F2 for filtering the composite signal to provide an F1 and F2 signal; a parameter estimator for determining an F1 and F2 period for the F1 and F2 signal for a sequence of frames and an average period shift (APS) for a current frame and a plurality of earlier frames; a metric calculator for calculating a metric that corresponds to the F1 and F2 period and the APS; and a decision processor for deciding whether the signal has been detected based in part on a comparison of the metric and a threshold.

Abstract: An improved adaptive line enhancer includes an adaptive Gray-Markel lattice notch filter having an adaptive notch frequency, in which the notch frequency is determined according to a notch frequency variable k. The value of k for the n+1th sample period is determined according to the following equation: k(n+1)=k(n)?sgn[y(n)]sgn[UPDATEFN]×? in which y(n) is the notch filter output, ? is the adaptation constant, and UPDATEFN has a transfer function in the z-transform domain of: ( ? - 1 ) ? ( k ? ( n ) - 1 ) ? z - 1 1 + k ? ( n ) ? ( 1 + ? ) ? z - 1 + ? ? ? ? z - 2 in which ? determines the bandwidth and k(n) is a variable for determining the current notch frequency. The algorithm for adapting the notch frequency enables the notch frequency to be directly calculated from knowledge of internal variables of the wave digital filter.

Abstract: A technique for performance-monitoring of a standard SONET signal involves first converting the optical signal to an electrical signal, removing in the framing signal time slot the framing signal for leaving only the framing signal noise in such time slot, and separating the framing signal noise from the data signal for viewing as a measure of the quality of the SONET signal.

Abstract: A system and method for determining DTMF coefficients stores a plurality of reference templates, at least some of said reference templates representative of likelihood ratios. A plurality of LPC coefficients is received. The LPC coefficients are representative of an input signal. The system and method determines a plurality of current likelihood ratios based, at least in part, upon the LPC coefficients. The system and method also determines an initial tone based, at least in part, on the minimum of the current likelihood ratios and the minimum of the reference likelihood ratios. A plurality of LSF coefficients is received. The LSF coefficients are determined, at least in part, upon the input signal. The system and method verifies the validity of the initial tone based, at least in part, upon the LSF coefficients.

Abstract: A general purpose network tone detection method and apparatus that allows the precise and accurate recognition of North American tones (MF, DTMF (Dual-Tone Multifrequency), and CPT (Call Progress Tones)) and international MF-R2 tones as well as taking into consideration other common tones such as Calling Card Service Prompt and Recall Dial. Through the use of the Discrete Fourier Transform (DFT) on small time windows and by providing phase continuity between these windows, the results of the successive DFTs may be combined and processed by a second DFT computation. This second DFT allows higher frequency resolution without requiring the re-computation of the DFT from the time samples. The resulting effect is a tone receiver with both high time and frequency resolution which consequently leads to robust and accurate tone recognition systems conforming even to the most stringent specification while maintaining low computational requirements.

Abstract: An apparatus for determining energy threshold detection values for detecting at least one tone having a known frequency and duration in an input signal. The input signal is input over a period of time that is divided into frame portions including at least an initial frame portion and a last frame portion. An energy value is determined for the at least one tone during each frame portion. A set of control flag signals are set for the at least two frame portions based on the energy values being below a threshold value in previous frame portions and the presence of noise in previous frame portions. An offset into a data table is determined based on the control word, the offset being used to retrieve a scaling coefficient and an address of an energy value for the at least two frame portions from the data table.

Abstract: An apparatus and a method for detecting the energy of a tone signal. A pre-process unit takes absolute values of transmitted samples and outputs the values to a delay unit for delaying the transmitted samples according to a predetermined delay interval. An operation unit subtracts the value calculated by multiplying the magnitudes of the samples transmitted before and after the time of the burst detection from the square value of the magnitude of the sample at the time of burst detection among the delayed transmitted samples. A detection unit detects the burst by confirming whether peaks corresponding to the number of the samples transmitted before the sample of the burst detection are detected by a signal of an operation value input from the operation unit. Therefore, without increasing device complexity, the probability of a false alarm caused by an impulse error and a number of burst errors is reduced; also stable burst detection can be performed.

Abstract: A tone detector includes at least one circuit (hereinafter “single phase reference matcher”) that not only performs a convolution of an input signal with a reference signal, but also compares the result of convolution with a threshold to determine if there is a match, and if so drives a signal active indicating that tone is present. If there is no match, another circuit (hereinafter “phase shifter”) delays the reference signal by a fraction (e.g. ⅛) of the measuring period, thereby to introduce a phase shift (e.g. &pgr;/8) between the input signal and the reference signal. The “single phase reference matcher” again performs the just-described operation, this time with a delayed reference signal, and repeats the operation as often as necessary (e.g. eight times) to cycle through the entire measuring period, thereby to ensure that tone (if present in the input signal) is detected irrespective of phase, during one of the operations.

Abstract: Improved techniques for time domain equalization are disclosed. The improved techniques include (i) improved time domain equalization techniques referred to as poly-path time domain equalization techniques; (ii) improved training methods for training transmitters and/or receivers of a data transmission system; and (iii) techniques for providing time domain equalization to a transmitter side of a data transmission system. These techniques are particularly suitable for time domain equalization in multicarrier modulation systems where channel shortening provided by time domain equalization is particularly needed.

Abstract: A decimating time discrete filter comprises transform means (2) for transfornming a real valued input signal i into a complex signal u represented by its real part Re{U} and its imaginary part Im{u}. The output signal u is down sampled by means of a decimator (4) which reduces the sample frequency by a factor of 2. Due to the downsampling process, a desired tone input signal can be translated to a different frequency. In order to prevent interference from signals at the input having the same frequency as the desired input signal, the transform means (2) have a transfer function with an increased width transition band from the frequency of the desired input signal to the frequency to which said input frequency is converted.

Abstract: A moving averaging filter which does not propagate a calculation error is provided reducing the size of the hardware. This moving average filter has a data holding unit for holding multiple successive data, a coefficient storing unit for storing coefficients, a first adder which calculates the sum of a pair of data of a prescribed combination held in the data holding unit, a multiplier which multiplies the sum to coefficient data obtained from the coefficient storing unit, and a second adder which adds up-a prescribed number of multiplication results produced by the multiplier.

Abstract: A system for wireless communication between a base station 30 and one or more remote stations 32 and 34 wherein a desired signal has associated therewith an identifier tone (a SAT in the vernacular of the AMPS system) and wherein interfering signals may have associated therewith identifier tones at different frequencies. A superresolution technique is used in the system to process the received data and to determine the relative magnitudes of any identifier tones which may be present in the received data. In the disclosed embodiment, the superresolution technique utilized is a least square error process. The resulting estimates are used by the system to facilitate the communication function.

Abstract: Supervisory audio tone (SAT) in a receive signal in a wireless communications system is detected using a discrete Fourier transform (DFT) supplied with samples of a filtered and decimated signal derived from the receive signal. The DFT can comprise M 2-input butterfly computation stages for computing an N-point DFT where N=2M. Each of the stages m from 1 to M−1 determines and stores, for each current sample at the input, at most 2m intermediate results forming only a subset of intermediate results required for computation by the next stage. Each of up to 2m other intermediate results required for computation by the next stage is produced by shifting of a respective stored intermediate result of the respective subset determined in respect of an input sample preceding the current sample by 2M−m−1 samples. For example there are M=5 stages for a 32-point DFT having 18 outputs, 6 outputs for each of three SAT frequencies.

Abstract: Algorithms for quickly and efficiently detecting a periodic signal such as a DTMF signal in a telephone system are described. The algorithms, implemented in a tone detector, detect the periodicity of the input data and do not require an estimation of the tone frequency.

Abstract: A signal detector for detecting a narrow-band signal of a desired frequency, using an adaptive type FIR filter having two filter coefficients, is disclosed. Using this filter, the input signal of a shift register which delays the signal, is estimated by referring to the output therefrom. A narrow-band signal of a desired frequency is detected by comparing each filter coefficient of this filter with a known filter coefficient corresponding to the frequency which is the object of detection.

Abstract: A level detector detects an input signal level. A rectifier (210) receives the input signal and provides a rectified signal. A prefilter (220) receives the rectified signal and attenuates high frequency components at frequencies near multiples of a decimation sample rate. The prefiltered signal is decimated (230) and low pass filtered by a lowpass filter (240) having a passband below the input frequency of the input signal. The level detector can be provided to control a variable gain stage circuit (935, 1010) which applies a gain to the input signal based on the level to form a dynamic range compressor or expander.

Abstract: An interpolation circuit for a digital filter which is small in circuit scale, operates at a high speed and is low in power consumption. The digital interpolation filter circuit includes front-end circuit 1 which outputs added value .SIGMA..sub.i of input data and the last data, and filter unit 2. Filter unit 2 includes delay circuit unit 3 which delays added value .SIGMA..sub.i of one clock and two clocks intervals, and outputs the delayed values, bit shift circuit unit 4 which inverts the signs of the added value and data obtained by delaying the added value 2 clocks interval and outputs resulting values and which shifts the data delayed one clock interval from the added value, 3 bits and one bit making multiplication of the delayed added value by 8 and 2, respectively, and outputs resulting values, adder 5 for adding the outputs of bit shift circuit unit 4 for each data having the same delay amount, and bit shift circuit 6 for shifting a result of the addition by 4 bits to divide the addition result by 16.

Abstract: A method of digital filtering and a digital filter having a filter response with a predetermined filter order. The filter response produces a predetermined set of filtered output samples defining an output signal from a received input signal. The difference between the power associated with the input signal and the power associated with the output signal is calculated. The filter order of the filter response is varied based on the calculated difference of power in order to change the filter response.

Abstract: The concept of integer ratio sample rate conversion is extended to fractional values, where the exact ratio is expressed as ?M+N/D!/L with L being a constant. This methodology allows sample rate conversion for all synchronous rates, where all rates are derived from the same crystal oscillator. For the implementation, the value of N/D is accumulated and depending on the overflow condition, either M or M+1 is selected as sample rate ratio for the conversion of each sample.