Spatio-Temporal Equalizer for a Receiving-Antenna Feed Array

Suppression of multipath effects and robust pointing would be achieved.

A spatio-temporal equalizer has been conceived as an improved means of suppressing multipath effects in the reception of aeronautical telemetry signals, and may be adaptable to radar and aeronautical communication applications as well. This equalizer would be an integral part of a system that would also include a seven-element planar array of receiving feed horns centered at the focal point of a paraboloidal antenna that would be nominally aimed at or near the aircraft that would be the source of the signal that one seeks to receive (see Figure 1). This spatio-temporal equalizer would consist mostly of a bank of seven adaptive finite-impulse-response (FIR) filters — one for each element in the array — and the outputs of the filters would be summed (see Figure 2). The combination of the spatial diversity of the feed-horn array and the temporal diversity of the filter bank would afford better multipath-suppression performance than is achievable by means of temporal equalization alone.

The seven-element feed array would supplant the single feed horn used in a conventional paraboloidal ground telemetry-receiving antenna. The radio-frequency telemetry signals received by the seven elements of the array would be digitized, converted to complex baseband form, and sent to the FIR filter bank, which would adapt itself in real time to enable reception of telemetry at a low bit error rate, even in the presence of multipath of the type found at many flight test ranges.

Each channel (comprising the signal-processing chain for a receiving feed horn) would contain an N-stage FIR filter. The incoming complex baseband signal in the ith channel at the nth sampling instant is denoted by yi(n). A filter weight at that instant is denoted generally by wi,j(n), where i is the index number of the channel (1 ≤ i ≤ 7) and j is the index number of the filter stage (0 ≤ j ≤N – 1). The signal-combining operation at the summation (output) point of the FIR filter bank is given by

where w_i,0≡1. The weights would be adapted by an algorithm known in the art as the constant-modulus algorithm, embodied in the following equation:

w_i,j(n + 1)= w_i,j(n)+ αy_i(n – j)z*(n)
[1 – |z(n)|²],

where α is an adaptation rate parameter.

In addition, the combination of the array and the filter bank would make it possible to extract, in real time, pointing information that could be used to identify both the main beam traveling directly from the target aircraft and the beam that reaches the antenna after reflection from the ground: Information on the relative amplitudes and phases of the incoming signals, which information would be indicative of the difference between the antenna pointing direction and the actual directions of the direct and reflected beams, would be contained in the adaptive FIR weights. This information would be fed to a pointing estimator, which would generate instantaneous estimates of the difference between the antenna-pointing and target directions. The time series of these estimates would be sent to a set of Kalman filters, which would perform smoothing and prediction of the time series and extract velocity and acceleration estimates from the time series. The outputs of the Kalman filters would be sent to a unit that would control the pointing of the antenna, enabling robust pointing even in the presence of multipath.