The Maximum Entropy Method (MEM) is compared to the periodogram method (DFT) for the estimation of line spectra given an error-free autocorrelation function (ACF). In one computer simulation run, a 250 lag ACF was generated as the sum of 63 cosinusoids with given amplitudes, Ai, and wave numbers, fi. The wave numbers cover a band from 0 to 89.239 cm-1with a Nyquist of 90 cm-1and are given to an accuracy of six significant figures. The amplitudes range over four orders of magnitude and are given to five significant figures. The resulting normalized ACF was truncated to five significant figures. The resulting MEM spectrum was remarkably accurate. The lowest 53 spectral peaks reproduced the input frequencies exactly (to six significant figures) while the largest error in the remaining 10 lines occurred at line 63: input = 89.2814; MEM = 89.2934. The relative amplitude of this tiny line is only 2×10-4and yet its frequency was recovered to one part in 9000.

Ionospheric scintillation effects encountered in the equatorial anomaly crest, polar cap, and auroral regions have been contrasted to provide information for the design and evaluation of the performance of multifrequency satellite communication links in these regions. The equatorial anomaly region is identified as the most disturbed irregularity environment where the amplitude and phase structures of 250 MHz andL-band scintillations are primarily dictated by the strength of scattering rather than ionospheric motion. In the anomaly region, the spectra of intense amplitude scintillations at these frequencies are characterized by uniform power spectral density from the lowest frequency (10 MHz) to 4 Hz at 257 MHz and to 1 Hz atL-band (1541 MHz) and steep rolloff at higher fluctuation frequencies with power law indexes of -5 to -7. Such structures are compatible with intensity decorrelation times of 0.1 and 0.3 s at 257 and 1541 MHz, respectively. The phase spectra at 244 MHz are described by power law variation of psd with frequency with typical spectral indexes of -2.4. The strong scattering at VHF induces extreme phase rates of 200° in 0.1 s. The 90th percentile values of rms phase deviation at 244 MHz with 100 s detrend are found to be 16 rad in the early evening hours, whereas amplitude scintillation can cover the entire dynamic range of 30 dB not only in the 250 MHz band but atL-band as well. In the polar cap, the 50th and 90th percentile values of rms phase deviation at 250 MHz for 82 s detrend are 3 and 12 rad, respectively, with comparable values being obtained in the auroral oval. The corresponding values for the S4index of scintillation are 0.5 and 0.8 in the polar cap, which are slightly higher than those recorded in the auroral oval. The power law index of phase scintillation at high latitudes is in the vicinity of -2.3, which is not a result of very strong turbulence as in the equatorial region but is considered to be a consequence of shallow irregularity spectral indexes. The phase rates at auroral locations are an order of magnitude smaller than in the equatorial region and attain values of 100° in 0.5 s. The extreme variability of ionospheric motion in the auroral oval sensitively controls the structure of scintillations. The long-term morphology (period 1979-1984) of intensity scintillations at 250 MHz in the polar cap shows that, in addition to the absence of diurnal variation of scintillations, and the presence of an annual variation with a pronounced minimum during local summer, there exists a marked solar control of scintillation activity such that it abruptly decreases when the solar activity falls below a threshold level.

A new model-order determination method for maximum entropy or autoregressive power spectra is presented. The method is based upon a simple F- (variance ratio) test for the reflection coefficients. The method is applicable to Burg-MEM, non-linear-MEM and unrestricted least squares methods and their many variants as well as to spectra derived via maximum entropy from autocorrelation estimates or measurements. The method will be tested on simulated signals - both wide-band: power-law processes and narrow band: sinusoids in noise.