| This document discusses the measurements conducted at Virginia Tech, in Blacksburg, VA, and at COMSAT Labs, in Clarksburg MD, to determine the impact of multipath fading upon a circularly polarized L-Band satellite-to-handheld communications system using an omni-directional antenna. A significant finding was that motion of the satellite could induce fades of up to 10dB, well in excess of the small link margins available to satellite-to-handheld systems (typically ≤3dB). A method for ameliorating multipath fading was then sought.; It was also found that the multipath fading effect was significantly different at as small a distance as half a wavelength from the original antenna position. Multipath signals could actually combine at the new position to form a small increase in signal to noise ratio (S/N), rather than a fade of as much as 16dB. Therefore it was determined that a small-scale antenna diversity system might offer a solution to the problem of multipath fading.; A series of experiments was set up with several different antenna types spaced roughly λ/2 apart to determine which orientation would provide the best S/N improvement. It was concluded that two vertically polarized patch antennas spaced vertically in the same plane by ≈λ/2 offered the best antenna space diversity method. Decorrelation between array elements was high, with multipath induced fades of up to 16dB experienced at one element being counteracted by a multipath induced peak signal in the other. A method was needed to determine the best way of using the signal improvements: either switching between antennas, or optimally combining the signals from both to get the best overall S/N.; It was determined that optimally combining the signals from each antenna, so that they matched perfectly in phase, yielded the best diversity gain, between 2 and 3 dB better than switching between antennas. |
