SATCOM-On-The-Move Antenna Tracking with Monopulse
By David E. Reed and William S. Sward • RT Logic, Inc.

SATCOM-On-The-Move (SOTM) terminals are challenged by antenna pointing and tracking requirements that are significantly more difficult than those for fixed satellite terminals. SOTM terminals have strict constraints on antenna size, weight and power [1]. Concurrently, the antenna tracking systems on mobile terminals become complex control systems due to the incorporated mechanical shock-isolation systems [2]. Figure 1 shows a compact SOTM antenna [3].

A monopulse technique is described as a method for improving the pointing and tracking accuracy of mobile satellite antennas with minimal increases in mechanical complexity. The monopulse technique provides a significant advantage in signal-to-noise ratio and, hence, Bit Error Rate (BER) performance for SOTM systems. Improved BER performance and fewer signal outages deliver robust link availability for the SOTM warfighter.

For harsh applications, such as military tracked vehicles, shock isolation techniques are often employed to mechanically protect the antenna system from the shocks and vibrations associated with a mobile terminal [2]. Figure 2 shows a shock-mounted antenna with wire-rope isolators. Unfortunately, the mechanical flexibility introduced by the shock isolation system complicates the antenna control system and can degrade pointing and tracking performance.

The monopulse antenna feed gives the radio system multiple antenna beam patterns to enable pointing. A typical system has multiple antenna feeds, two to five feeds depending on the performance required. A two-feed system typical uses a main beam and difference beam. The difference beam feed is typically time multiplexed between the azimuth and elevation dimensions. Performance is improved with additional, dedicated feed beams and more receive channels. There are numerous tradeoffs between the number of beams, number of receive channels, SOTM performance, size, weight and cost.

For the purpose of illustration, a five-beam monopulse SOTM system is considered here. One center-fed beam provides the best beam for direct line-of-sight transmit and receive. Two offset feeds on either side of center provide direction discrimination in one axis (e.g., azimuth), while another two beams do the same in the orthogonal axis (e.g., elevation).

Figure 3 shows example patterns for the azimuth axis. In this example, the half-power beam width is approximately 3.6 degrees and each offset beam is spaced by qo = 2 degrees from center. Figure 4 plots typical antenna gain patterns for each of the three beams.

The antenna pointing/tracking control system uses the offset beams to estimate the pointing error and drive this error toward zero. Figure 5 shows a block diagram of a traditional single-beam antenna pointing control system.

With the conventional single-beam approach, the antenna tracking loop operates on the received signal strength of the satellite downlink. While many satellites provide a downlink beacon for these purposes, it might be advantageous to track on the downlink communication traffic. Unfortunately, a single-beam approach does not provide direction information for the angle-off-boresight. Hence, a scanning technique is typically required to determine the azimuth and elevation errors. Scanning techniques, however, reduce the average received signal power, which directly degrades the received BER performance. Additionally, scanning results in angular error measurements that are not strictly correlated in time and the scanning process requires significant time to generate the errors. These limitations necessitate a narrower tracking loop bandwidth to ensure control loop stability. Consequently, the narrower bandwidth limits the loop’s ability to track rapid dynamics due to shock and vibration.

The monopulse tracking approach is similar in architecture but employs additional error tracking signals simultaneously, as shown in Figure 6.

The spacing of the beams relative to the beam width determines the directional gain of the pointing angle error signal. Figure 7 shows the error signal versus pointing error for spacing offsets of qo = 1°, 2°, and 3°. The larger spacing has higher gain (or sensitivity), but a smaller linear range, as evidenced by Figure 7.

The slope of the error signal, for small errors, is a gain factor in the tracking control loop. Given a nominal noise level and disturbance that perturb the measurement, a higher gain or sensitivity generally leads to more accurate pointing. The slope of the linear portion of the error curve is plotted in Figure 8 versus the offset angle. This quantifies the loop gain when stable or for small errors.

Without the monopulse feed, this error signal would be generated by scanning the antenna, i.e., pointing it to either side of the estimated LOS in order to measure signal amplitude, as shown in Figure 8. Scanning in both axes is often accomplished by a conical scan where the antenna rotates around the LOS.

The monopulse feed has the advantage over the scanning technique because it allows both offset measurements to be taken without pointing the main beam away from LOS. Also, the measurements are taken at exactly the same time so that many error sources, such as fading, multi-path or vibration are perfectly correlated between the two. Therefore, the monopulse error signal is robust against a variety of impairments found in SOTM applications. The following section considers a typical impairment and quantifies the communication performance improvement of the monopulse technique over a conventional single beam system.

All antenna pointing control systems are limited in accuracy and bandwidth, and generally have a residual pointing error when disturbed by mechanical motion. Often, this error is induced by platform vibration and vehicle dynamics. This section considers an example disturbance and quantifies the performance improvement of a monopulse system.

The residual pointing error is treated as an undesired disturbance added to the tracking loop filter output. This error is tracked out by the loop to the extent that it is within the control loop bandwidth. However, any high frequency portion of the disturbance remains as pointing jitter and results in amplitude modulation of the received (and transmitted) signal.

Figure 9 shows an example waveform representing a possible residual error from shock/vibration disturbance to the SOTM antenna system. The example perturbation has multiple frequency components and large spikes in the time domain.

Because the monopulse feed has additional beams pointing slightly off-axis with respect to the center of the main beam, it provides antenna gain over a wider angular range than a traditional single beam system.

As the example disturbance of Figure 9 is applied to the monopulse tracking system, each antenna beam experiences a different amplitude modulation based on the magnitude of the angle off-boresight. Figure 10 shows the gain of the three monopulse beams (of the azimuth axis) with the residual pointing error waveform. A complete model would include perturbations in both the azimuth and elevation axes. In Figure 10, the residual antenna pointing error was scaled for peak pointing error of 3°; the azimuth offset beams are black and red, while the main center beam is green.

The received signal for the single-beam system has amplitude fluctuation imparted by the antenna gain variation:

As(t) = A0(t) = G(q(t))

Hence, the green trace of Figure 10 is representative of a traditional single beam system. As observed, there are significant reductions in the effective antenna gain as the pointing errors deviate from boresight (black trace of Figure 11). Conversely, with all three beam signals available to the receiver, the largest amplitude signal is selected in the monopulse implementation. This results in smaller signal amplitude reductions due to the residual pointing error (red trace of Figure 11).

The simple monopulse beam processing effectively increases the minimum antenna gain, which has a positive effect on the received BER. In this case the received signal amplitude is:

Am(t) = max{ A0(t), A1(t), A2(t) }

Over a period of time, such as multiple periods of the vibration waveform, the receiver performance is dominated by the minimum signal level, i.e., min{As(t)} for the single beam case and min{Am(t)} for the monopulse beam case.

Figure 12 shows the minimum antenna gain versus the peak residual pointing error for our example. Notice that small peak pointing error results in no advantage for the beam selection processor. However, larger pointing errors show significant gain advantage as the center beam gain rollsoff.

The monopulse technique mitigates the antenna pointing errors associated with mobile SOTM terminals. Compared with traditional scanning and lobing implementations, the monopulse technique provides improved effective antenna gain with residual pointing errors. Antenna scanning and lobing techniques rely upon antenna motion to determine the pointing errors driving the control loop. Dynamic interaction between intentional antenna movement (scanning) and motion due to vehicle movement create a complicated environment for a control system. The monopulse technique with a multi-channel receiver avoids the time varying errors (modulation) associated with antenna scanning and lobing techniques. The monopulse approach provides time-coincident angle-off-bore-sight measurements in both azimuth and elevation dimensions.

REFERENCES
[1] V. Weerackody and L. Gonzalez, “Mobile Small Aperture Satellite Terminals for Military Communications,” in IEEE Communications Magazine, Vol. 45, No. 10, October 2007, pp. 70-75.
[2] J. DeBruin, “Shock Isolation for Mobile Pointed SATCOM Systems,” MILCOM 2008.
[3] J. DeBruin, “Control Systems for Mobile Satcom Antennas,” in IEEE Control Systems Magazine, February 2008, pp. 86-101.

William S. Sward is the Engineering Manager for RT Logic, Inc. William S. Sward received a BS in Electrical Engineering from Iowa State University and an MS in Electrical Engineering with emphasis in RF Communications from the University of Colorado. He is currently an Engineering Manager at RT Logic working on RF data links, satellite communications and radar. His previous experience includes working with commercial wireless modems, fading channel emulators, CATV modems, GPS receivers, and radar systems for various corporations. His current interests are RF communication systems based on DSP platforms. He can be reached at bsward@rtlogic.com.

Click here for reuse options!
Copyright 2010 Webcom Communications Corp.