Search

This paper analyzes the performance of digital phase modulation systems in Gaussian noise and determines required signal-to-noise ratio as a function of the number of discrete phases and the desired error rate, under conditions of no fading. Both coherent detection with a locally-derived reference carrier and phase comparison detection are considered. The calculations show that multiphase modulation provides an efficient trade of bandwidth for signal-to-noise ratio in comparison with multilevel amplitude modulation. It is also found that phase comparison detection introduces about a 3-db degradation over coherent detection except with binary modulation, for which the degradation is less than 1 db for error rates not exceeding about 0.001.

Frequency translation of an RF wave can be accomplished by subjecting it to a constant rate of phase shift or by applying an ideal repetitive sawtooth phase function which snaps to zero at 360 degrees (serrodyne modulation). The DIGILATOR is a new type of frequeucy translator which employs a multibit latching ferrite phase shifter with a suitable digital switching driver to obtain a multiple-step approximation to the serrodyne phase function. This paper reviews the theory of an N-step DIGILATOR and describes the design of a 16-step, X-band device. Experiments showed that performance is independent of microwave frequency over the design frequency band of the phase shifter. The carrier and close-in sidebands were suppressed by 39 dB and predicted far-out sidebands were within 1 dB of theoretical amplitudes.

The need for digital simulation or modeling of linear modulated communication and control systems frequently arises. This paper presents a method for accomplishing this task which has outstanding computational efficiency. It applies to those situations in which the input to the modulator or detector can be presented as a sum of sinusoids of arbitrary amplitudes, phases, and frequencies. Post modulator or post detector filtering is assumed to be accomplished by a linear stationary differential system. The method is considerably more efficient than conventional methods because the action of the modulator or detector is accounted for analytically and because an alternate expression for the transition matrix is employed which eliminates derivatives of the input signal from the integrand of the convolution of the transition matrix. The advantages accruing to these measures are demonstrated by two examples, one of which is frequently encountered in discrete phaselock loops and in identification of system parameters by sinusoidal response methods.

In this paper one establishes: (a) equations for the mass transfer coefficient in the continuous phase from a single spherical bubble or drop, valid for the whole range of values of the tangential component of velocity at the interface; (b) equations for the mass transfer coefficient in the continuous phase from a bubble which belongs to a group of bubbles. The answer to the first problem is obtained for and using a simple interpolation method. A comparison between the results given by this method and the integral method is made. The second problem is investigated by means of a cellular model.

Multi-h continuous phase modulation (CPM) employing a constant envelope modulation technique is being used over UHF military satellite communication (MILSATCOM) channels. It was implemented as a result of the Advanced Digital Waveform (ADW) program. The goal of this program was to increase throughput on UHF SATCOM channels. The resulting waveform, described in MIL-STD-188-181B, is a 4-ary, 2-h, linear shaped, full response CPM. This paper describes techniques for further improving the power and bandwidth performance of multi-h CPM. Improving power efficiency will benefit disadvantaged platforms by allowing link closure at lower C/N/sub 0/ levels. Improving bandwidth efficiency will allow higher data rates to be used. Three methods for accomplishing these goals are explored. The first is to increase the symbol alphabet size from 4 to 8 symbols by encoding 3 bits per symbol instead of 2, thus increasing bandwidth efficiency by 50 percent. The second area of interest is to implement a partial response waveform, meaning that each symbol is shaped over multiple symbol periods. This second method will add more memory to the trellis structure and should also cause the spectrum to become more compact. The third method investigated is to increase the number of modulation indices. The idea is that using three or more indices will spread the trellis more, increasing the minimum Euclidean distance. The first two areas of research, 8-ary CPM and partial response CPM, show promising results, in some cases improving both power and bandwidth efficiency. Of these two, 8-ary CPM seems the more practical path for immediate further development. Partial response is actually more promising in theory, but is much more complex and requires further research. Early results of three-h CPM show that adding another h-value does not improve performance and does not appear to be a promising avenue for further research.