Abstract: A reduced cost digital data modulation and demodulation scheme for very small aperture terminal (VSAT) satellite communication systems combines non-coherent frequency detection with trellis-coded, multi-frequency modulation. The technique is robust to phase noise and frequency uncertainty and is capable of achieving the performance of rate one-half phase shift keyed modulation. The code width of data to be transmitted bears a prescribed relationship to the size of a multiple frequency set and the selection of a given combination of frequencies within that set for transmission during a respective baud. One portion the data is convolutionally encoded and provides a pointer to one of plural groups of orthogonal signal sets associated with frequencies into which the multiple frequency set has been partitioned. Another portion of the data identifies the frequency combination within the group pointed to by the one portion. The combined portions encode a multi-frequency waveform transmitted to the receiver site.

Abstract: A reduced cost, robust data communication scheme for very small aperture terminal satellite communication systems combines non-coherent frequency detection and trellis-coded, multi-frequency modulation. The data code width is associated with the size of a multiple frequency set and the selection of a given combination of frequencies for transmission during a respective baud. One portion the data is convolutionally encoded and points to an orthogonal signal set associated with frequencies of a partitioned multiple frequency set. Another portion of the data identifies the frequency combination within the group pointed to by the one portion. The combined portions encode a multi-frequency transmission waveform. At the receiver, the multi-frequency tone sequence is detected by non-coherent frequency detection. For each baud, matched filters output a most likely set of frequencies transmitted during that baud, as soft decisions to a Viterbi decoder.

Abstract: An adaptive bit synchronizer is operable to extract digital data and its associated clock from a transmitted digital signal, and includes a tunable matched filter set for modifying the input signal to correct for deviations in offset and gain, which filter set includes data, transition and derivative matched filters. A sampling device samples the output of the data matched filter for making bit decision and for estimating the reliability thereof. A clock-producing device is connected with the matched filter set for producing at least two clocks, use being made of an optimum phase detector for estimating the time error between the proper clock edge and the actual clock edges. A loop filter circuit smooths the estimates of the proper clock time to generate clock signals, and a device responsive to the average square error of the clock signals varies the loop parameters of the loop filter means to minimize average square phase error.

Abstract: A system and method are disclosed for synchronizning the linear PN sequences contained in a received spread spectrum signal, characterized by the provision of a resident PN generator that is responsive to the chip rate clock for producing a replica of the PN sequence with arbitrary phase, a running matrix inverse of the matrix (R) formed by n successive observations of the register of the resident generator, and a matrix vector product device for multiplying the running inverse by a column vector of noisy chips, thereby to obtain a plurality of estimates of the phase vector. These estimates are smoothed and averaged to produce the smoothed phase vector (c.sub.j) that is applied to one input of a dot product device that operates in conjunction with the contents of the shift register of the resident generator to produce the properly phased PN sequence, which sequence is then supplied to despreading means for combining the noisy chips with the properly phased PN sequence.

Abstract: A method of transmitting binary data from one station to another via a troposcatter medium, characterized in that the data is converted to parallel form so that the bits produce distinctive pairs of sine and cosine harmonics having different frequencies, which harmonics are summed in two separated channels that are modulated by rf sine and cosine modulating signals that are combined and transmitted to the receiver, together with a test signal that was periodically inserted in the parallel bits. The receiver supplies the signals to banks of matched filters that produce a first set of signal estimates from which the test signal is detected. A matrix system responsive to the test signal produces from the first set of estimates a second set of signal estimates having lower distortion than the first set.

Abstract: A method of developing waveforms having the energy therein concentrated within a limited bandwidth and modems for generating, transmitting, and receiving digital data utilizing such waveforms as carriers. A set of mutually orthogonal basis signals are selected to maximize their energies within the channel, each is binary coded and stored in the modem. A set of optimized coefficients is stored and utilized to weight each basis function in a coding arrangement to define bytes of digital data. The coded and weighted basis functions for each successive byte are summed to form a composite signal and transmitted over the channel as an analog waveform. Stored basis signals at the receiving end are correlated with the received waveform to thereby extract the coefficients which are decoded to reproduce the transmitted bytes. Both coding gain and reduction of intersymbol interference is achieved.