Multipath and diversity studies in meteor scatter propagation.

Meteor scatter (MS) multipath and diversity is investigated using 37 MHz continuous wave transmissions. A cross-correlation technique and a modelled MS communications system are used to analyse diversity-data collected at temperate latitudes at different antenna separations and season. Cross-correlation studies are first performed at an antenna separation of 102 by investigating the crosscorrelation variation, with time, of signals which have been categorized as "underdense," "overdense," or "not known" (NK). The cross-correlation coefficients of signals from underdense and overdense trails are shown to be high when correlated over their total signal-envelopes. This is not true, however, when the cross-correlations are performed on signal envelopes which have been segmented in time. NK signals are observed to be more decorrelated than underdense or overdense signals and, therefore, are likely to be the most advantageous to a MS communications system incorporating space-diversity. Combining the data from all three signal categories it is shown that ~40% of signals exhibiting a duration 0.75s, have correlation values of less than 0.6 after the first 0.25s of signal decay. The correlation-time dependency observed for NK and underdense signals is not identified for the overdense signal category. It is proposed that for underdense and NK signals the correlation-time dependency is due to the vector addition of other weak signal-modes. These weak signal-modes, however, have little effect on overdense signals which, typically, exhibit higher signal-powers during the early stages of signal decay. Cross correlation analysis of signals received at antennas separated by 5°, 10°, and 20° shows that no spatial variation, and in particular no decrease, in average cross-correlation coefficient is observed for underdense or NK signals. At each antenna separation the cross-correlation coefficients of these two signal categories were strongly dependant on time. Overdense signals, however, show no cross-correlation time-dependency at 5° and 10°, but a strong time-dependency is observed at 20°. The measurements support the view of previous workers who have suggested that a 4° antenna separation may be useful in a MS diversity communications system. At an antenna separation of 10? a modelled broadcast scan-diversity and maximal-ratio diversity systems are investigated. The results show that the optimum broadcast data-block duration, in February, for a non-diversity system is ~90ms assuming a ~10ms preamble. The diversity-gain results, at a signal-to-noise (SNR) ratio of 3 dB in a 3 kHz bandwidth, show that the broadcast throughput scan-diversity gain is ~1.4 in June and ~1.18 in February. These gains are equivalent to transmitter power improvements of ~3 dB in June, and ~1.4 dB in February, assuming a simplified relationship between transmitter power and data throughput. The broadcast throughput gain achieved by a maximal-ratio diversity system in June is ~1.9 and ~1.6 in February. At the 3 dB SNR these gains are equivalent to transmitter power increases of ~5.8 dB and ~4.1 dB in June and February respectively. The summer-winter diversity gains observed are attributed to the reception of more decorrelated NK signals (e.g. sporadic-E) in June than February. Frequency shift keying (FSK) error-probability analysis is investigated for a selection-diversity system and a maximal-ratio diversity system with antennas separated by 10°. The results show that by using signals 3 dB SNR, ~62% fewer errors are observed in February using selection diversity than a non-diversity system, and that in June this improvement decreases to ~50%. Results obtained from the maximal-ratio diversity system show that in both February and June ~90% fewer errors are observed than a non-diversity system. By assuming 216 bit broadcast data-blocks, the results show that a non-diversity system, using signals 3 dB SNR in February, will experience an error every 8th data-block. A selection-diversity system and a maximal-ratio diversity system, however, will experience errors every 24th and 78th data-block respectively. In June every 6th, 16th and 52nd data-block will be received in error by the non-diversity, selection-diversity and maximal-ratio diversity systems respectively.