A geometric propagation model is developed for simulating the reception of direct sequence pseudonoise (DSPN) satellite signals by a directional or omni directional antenna in the dense scatterer mobile environment. The model is used to predict fading statistics. As expected, the mitigation of fading is closely related to the ratio of DSPN chip duration to delay spread of the scatterer medium for both the omni and directional antennas. The results suggest a general, frequency domain interpretation for the value of DSPN in mitigating multipath fading.>

Previous analyses are extended to give the performance of the adaptive 2 bit A/D converter in combined CW and Gaussian interference. The converter is seen to give substantial conversion gain even when the direct sequence pseudonoise (DSPN) spread signal is much weaker than the Gaussian component of interference. The upper bound on conversion gain depends primarily on the relative strengths of the Gaussian and CW components of interference. The relevant relationships are given in graphical form.

Considerable recent attention has been given to the use of pulse modification to reduce spectral sidelobes in constant envelope data transmission within the "offset" modulation format that includes offset QPSK and MSK. Analytic techniques have been developed which predict probabilistically (without dependence on simulation) the power spectral density that results from pulse modification combined with hard limiting where necessary to achieve constant envelope. Here the computational results are compared with experimental results. Agreement between theory and experiment is seen to be good. The implications for bandwidth efficiency and communication efficiency are discussed.

The non-Gaussian nature of CW interference is exploited to suppress the CW via the A/D converter following chip demodulation in a direct sequence pseudonoise (DSPN) communication link. The modulation is coherent BPSK. The A/D converter quantizes to 2 bits, sign and magnitude. A scheme of threshold adaptation and postquantization weighting gains great advantage from making very reliable decisions on a relatively small percentage of the demodulated chips. The bit error rate performance in CW interference generally surpasses that of an ideal analog DSPN correlator. The performance in Gaussian noise is within 0.6 dB of ideal analog.

A new class of basic pulses is introduced into the MSK signaling format to gain improvements in the out-of-band power spectral density characteristic. Some loss in communications efficiency is incurred, but the loss may in certain cases be outweighed by the spectral improvements that are made available.

Modulation and demodulation in the minimum (frequency)-shift keying (MSK) format are reduced to the form of coherent biphase keying. Considerable simplification of circuitry is thus effected without sacrifice of performance. When the ratio of the carrier frequency to the bit rate is high, then the technique described here does not require the precise relative phasing of any pair of oscillators in the transmitter. Experimental results demonstrate the feasibility of synthesizing the one filter that is unique to this signaling scheme. In the present instance, the filter is realized as a surface acoustic wave device.

The minimum (frequency) shift keying (MSK) format is generalized to enable further improvements in the spectrum while retaining the well-known constant envelope property and the excellent communications efficiency of MSK. A specific alternative to MSK is presented, and the spectrum is compared with that of MSK. The theoretical limitations on possible further spectrum improvements are discussed.

Knowing that the tails of the ideal\sin \pi t/ \pi tpulse represent a divergent series, the problem is stated of finding distortionless bandlimited waves which give rise to a minimum of the worst peaks of voltage that can appear on the channel in unit-height binary transmission. The optimization problem is then solved for all sampling rates up to1/2the Nyquist rate by the unique wave which never gives rise to more than one volt on the channel, just the magnitude of the data bits themselves. An upper bound is established on the worst channel voltage at higher rates by means of a class of pulses whose spectrum reduces to the optimum when the sampling rate parameter is chosen to be1/2the Nyquist rate. A time-domain optimization principle is stated which deals with zero manipulation, and the implications for the concept of dimensionality are discussed. The principal results of this paper are related to other system concepts, including timing-jitter immunity and duobinary transmission.

The minimum Gabor bandwidth ofMorthogonal real signals which together occupy a given time interval is derived. For special values ofM, one form of the optimum set ofMsignals is specified. Interrelations of the Gabor definition of bandwidth to other definitions are also discussed.