Adaptive Nulling Antennas for Military Communications

Don J. Hinshilwood and Robert B. Dybdal

Adaptive antennas dynamically respond to interference by "nulling out" or canceling the interference energy. The Aerospace Corporation has successfully applied the technology to satellites and ground terminals for two Air Force Space and Missile Systems Center programs.

Designers and operators of military communication satellites have long sought to reduce the vulnerability of their systems to intentional and unintentional interference. Their efforts have resulted in three distinct technologies for interference mitigation: spread-spectrum modulation, antenna sidelobe reduction, and adaptive interference cancellation.

An adaptive uplink antenna forms a null in the direction of a jamming signal, preventing it from interfering with communications to intended users.

Spread-spectrum modulation, the most widely used, modulates the communication signal using a code that is known to both the sender and receiver. This additional modulation expands the bandwidth of the transmitted signal, thereby spreading the interference over a wider frequency range and reducing its effect.

Antenna sidelobe reduction seeks to minimize interference received beyond the antenna's desired field of view by applying antenna design techniques to reduce sidelobes—the minor lobes separated from the antenna's main beam (see sidebar, Adaptive Sidelobe Cancellation). This technique effectively reduces interference both received and transmitted by the antenna.

Adaptive interference cancellation monitors the received signal and identifies interference when present. Several antenna elements—the actual radiating and receiving components—are combined to form a null or cancellation in the direction of the interference. The adaptive system automatically responds or adapts to changing interference patterns, without human intervention (see sidebar, Adaptive Nulling Antennas: The Next Wave in Jamproof Communications).

Practical system designs must employ an appropriate combination of these three techniques to achieve the desired level of interference protection in the most economical way (see sidebar, Adaptive Uplink Antennas: How Do They Work?).

Adaptive Antenna Designs

Adaptive antenna systems contain five major components: a means of detecting interference, a means of distinguishing desired signals from interference, a control processor for determining how to combine the antenna elements, antenna elements and circuitry to respond to commands from the control processor, and a performance monitor to identify changes in the interference environment and respond accordingly.

Development of adaptive systems began in the 1950s with two nearly concurrent design approaches. The first focused on sidelobe cancellation, an approach that assumes interference arrives through the antenna sidelobes while the desired signal arrives through the main beam. Auxiliary broad-coverage antenna elements are used to sample the interference. The correlation between the main and auxiliary antenna elements quantifies the received interference power, which is treated like an error signal in an automatic control system. Minimizing the error signal is equivalent to minimizing the interference.

This parametric plot illustrates the tolerances for the amplitude and phase matching in achieving a specified null depth. Maintaining a near-zero interference level is demanding, and the required tolerances become more stringent as the cancellation performance increases. For example, the requirements for a 40-decibel cancellation (1/10,000 of the interference power) are significantly tighter than those for a 30-decibel cancellation (1/1,000 of the interference power).

The second approach applies to array antennas—units equipped with several primary receiving or radiating elements; in this case, the array elements are combined under adaptive control to maintain desired signal reception and form nulls in the direction of interference sources.

This process of canceling interference is inherently a subtraction operation. The radiation patterns or spatial variations of the antenna elements are subtracted from one another to form an overall null in the direction of the interference.

It's a demanding process, and the formation of deep nulls requires extremely precise matching of amplitude and phase characteristics across the operating bandwidth. How precise the system can be depends on the frequency variation of the antenna elements and the amplitude-tracking and phase-tracking performance of the electronics. The level of design tolerance needed to maintain adequate precision poses a significant challenge for practical adaptive antennas.

Technological advances in recent years have helped. Low-noise system components, for example, are available with the necessary unit-to-unit reproducibility at a low weight with high performance. Adaptive weighting circuitry, which is needed to combine the antenna elements, provides the high accuracy needed to meet the stringent tolerances. Digital technology for the adaptive processing and control functions is well developed. Capable instrumentation is available to evaluate and diagnose adaptive system performance. Simulation techniques have also been developed to provide high-fidelity modeling of practical hardware and evaluate system performance on a statistical basis.

Satellite Uplink Antennas

Aerospace studies of multiple-beam antennas in the 1960s led to the Defense Satellite Communications System III satellite. These studies also spawned the idea of adaptively processing the individual beams in order to cancel interference from ground sources; in fact, they are believed to be the origin of adaptive uplink designs using multiple-beam antennas. Subsequent development by both Aerospace and the research community advanced the concept. Aerospace efforts in the Extremely High Frequency Follow-On program, for example, sought to develop a design with sufficient simplicity to allow practical implementation; these efforts focused on demonstrating the performance of a seven-beam design.

A flight unit of the Milstar 2 medium-data-rate adaptive nulling antenna. Behind the main reflector (the large disk at the bottom of the photo) sits the radio-frequency correlator, beamforming network, uplink-feed elements, and control electronics. The downlink feed and dichroic subreflector are positioned further above. Uplink user and interference signals are focused by the main reflector and subreflector onto the cluster of feed elements. Under the direction of the nulling processor (not shown), the beamforming network produces the nulls in the antenna's pattern. The correlator receives samples of the uplink signals from the beamformer, measures their strengths, and sends this information to the processor to finish the adaptive feedback loop. (© TRW Inc. 2000. All Rights Reserved. Republished by kind permission of TRW Inc.)

The results are apparent in Milstar MDR (medium data rate), a system that provides substantial interference protection. The design combines all three protection techniques for interference mitigation. Spread-spectrum modulation, for example, helps reduce interference susceptibility; the use of low-sidelobe multiple-beam antennas reduces interference received beyond the design coverage area; and adaptive antenna processing serves to protect users from interference within the design coverage area. The system's predecessor, Milstar LDR (low data rate), achieved adequate interference protection simply by combining low-sidelobe multiple-beam antennas and spread-spectrum modulation; however, spread-spectrum modulation becomes less effective at higher data rates (because of the decreased ratio of signal bandwidth to transmission bandwidth). Thus, for the MDR system, adaptively processed uplink antennas were needed to maintain a high level of interference protection.

This figure illustrates the difference between a standard (left) and an offset (right) reflector antenna. The feed of the offset antenna is directed toward the dish at an offset from its axis of symmetry.

The uplink antennas on the satellites use an offset reflector and a cluster of focal-field radiators to obtain multiple beams within a spot-beam coverage area. Seven beams are adaptively combined in this design. When interference is not present, a predetermined combination of the individual beams provides the quiescent pattern. The antenna is mechanically steered to its desired location and can be changed in orbit as communication needs vary. When interference begins, the individual beams within the cluster are adaptively combined to create pattern nulls to cancel the interference. Meanwhile, the low sidelobes provided by the multiple-beam design prevent interference beyond the coverage area. Interference can be detected in various ways. For example, any received signals that do not contain the proper spread-spectrum coding are treated as interference.

The satellite antenna is configured to communicate with a specified coverage area. When interference is present, however, the nulls created by the antenna system reduce the available coverage. The percentage of design coverage area still available during interference provides a useful metric for assessing the performance of candidate designs and specifying system performance (see sidebar, How Close Can the Interference Be?).

Protecting Ground Terminals

Aerospace also helped develop a sidelobe-cancellation system for the Defense Satellite Program (DSP) as a part of the Satellite Readout Station Upgrade program. Protecting ground receiving stations from nearby interference was the primary goal. The waveforms for this system were already well established, so spread-spectrum modulation could not be employed. Also, the large antennas at the ground terminals were enclosed in radomes—protective radiolucent shells for housing antennas—so sidelobe-reduction techniques could not be used, either. Therefore, the only feasible option was an adaptive sidelobe canceller.

A fundamental challenge in designing this system was the sidelobe response of the main antenna. The sidelobe response describes the sum of individual sidelobe contributions from various sources, including direct radiation from the feed, diffraction from the reflector rim, and scattering from feed supports. Each sidelobe source arises at a different physical location on the antenna and has a corresponding time delay. As a result, the sidelobe response of the reflector antenna varies with frequency over the operating bandwidth. This frequency dependence must be matched by the frequency dependence of the auxiliary antenna elements and adaptive circuitry. The greater the cancellation required, the tighter the tolerance in matching these frequency responses.

Shown here are representative patterns processed by a sidelobe canceller, which assumes interference arrives though the antenna sidelobes, not the main beam. By correlating the main and auxiliary antenna elements, the system can identify interference and automatically begin countermeasures.

At the ground receiving station, the frequency responses of the main reflector antenna and the auxiliary antennas differed greatly, so the adaptive circuitry needed to provide significant equalization to achieve effective cancellation over the required bandwidth. In this case, an adaptive transversal filter was successfully used. This filter combined time-delay components with adaptive weighting circuitry. These time-delay components were initially measured using a scale-model antenna and later verified with the full-scale antenna. The measurements were performed using a general-purpose vector network analyzer. The measured amplitude and phase response was then processed by the network analyzer, which converts measured frequency-domain data into time-domain data. This time-domain response of the antenna sidelobes provided the required time-delay spread for the transversal filter.

This sidelobe canceller subsystem has been installed at DSP ground sites and has been successful in protecting mission data from terrestrial interference.

Adaptive Antenna Simulation

Simulation analyses are an important part of adaptive antenna development. Initial simulations, based on a scenario that describes the anticipated interference environment, are used to develop a hardware design capable of complying with system requirements. As system development proceeds, actual hardware data can be incorporated into the simulation, thereby improving its usefulness.

adaptive performance with 2 simultaneous jammers

This pair of figures shows the performance of the adaptive antenna system as time passes after two jammers turn on simultaneously. The upper plot shows the drop in jammer-power-to-noise ratio as the system cancels the jamming. The lower plot shows the percent coverage area recovered for user communications. When jamming begins, percent coverage area falls but quickly recovers. These time variations illustrate the adaptive system's transient response—the time it takes for the system to reach a steady state after interference begins.

Simulation also plays an essential role in testing adaptive antennas. Although the simulation can be run repeatedly with varying scenario parameters to provide statistical answers for system performance questions, such testing can take a long time. A more reasonable approach is to use the simulation to derive representative test cases, which can then be measured to validate the simulation. Once validated, the simulation provides the required statistical answers to questions about system performance.

In fact, Aerospace developed a detailed simulation package, called NullerSim, to facilitate this process. Originally intended as an independent means of verifying contractor analyses, NullerSim proved useful in addressing system performance questions that arose during development. In the case of the Milstar MDR nuller, the simulation has been fully validated by measured performance and has furnished support for the on-orbit testing of the MDR design. More recently, NullerSim was used to derive requirements and provide validation and benchmark values for contractors' simulations for the Advanced Extremely High Frequency program.

Adaptive System Testing

In evaluating an antenna's ability to reject interference, the researcher must answer two fundamental questions: What is the steady-state loss in performance when interference is present, and how soon after the start of interference does the system reach steady-state performance? Answering these questions requires testing at the system level rather than at the component level.

In this pair of figures, the second jammer turns on after a delay relative to the first jammer. The percent coverage area falls when the first jammer turns on, but then recovers. It again falls when the second jammer turns on, and recovers once again.

Performance measures for adaptive antennas differ from those of conventional antennas. Adaptive operation is generally tested by measuring signal-fidelity parameters. For example, measuring the bit-error rate of the desired signals in the presence of interference provides a good metric of steady-state performance. By measuring the time required for the adaptive weighting values to reach their steady-state levels after the start of interference, the transient response of the adaptive system can be obtained. The ability to freeze the adaptive weight settings and measure the antenna pattern also provides useful information, such as percent coverage area for an uplink antenna or gain and system-noise temperature variations for a sidelobe canceller.

Evaluation of adaptive antennas also requires the ability to generate desired and interfering signals arriving from different directions. To test the satellite readout station for the DSP upgrade, interfering signals were obtained from small antennas located close to the ground terminal antenna. Evaluating the uplink adaptive antenna for the Milstar MDR program was more challenging because of the high risk involved in testing flight hardware outdoors. Aerospace developed a novel concept that not only solved the problem, but achieved significant cost and time savings as well. This concept extends so-called compact-range technology, a technique that provides test signals in an indoor environment.

A compact range normally uses the plane wave generated in the near field of an offset reflector whose diameter is at least twice as big as the antenna being tested. If the compact-range reflector is envisioned as an antenna, a subreflector can be used to provide high-fidelity beams off-axis from the focal point. This technique is commonly used to obtain the required field of view for uplink antennas, as indeed it was for the Milstar MDR design. When applied to the compact range, such a design results in plane-wave components arriving from different directions—as required to evaluate the uplink adaptive design. Discrete illuminators placed within the focal region simultaneously produce test fields having different arrival directions, while independent signal sources provide desired and interference signal components.

Conclusions

Further developments in adaptive antennas will improve communications in adverse environments. Development of associated technology in high-performance radio-frequency components, digital processing, and control electronics as well as increased understanding of simulation and testing techniques will lead to more capable and economical adaptive antenna designs. Research at Aerospace has already helped reduce the vulnerability of defense communications satellites (see sidebar, Research Endeavors). Future military and commercial systems—particularly in the area of wireless communications—can look forward to similar improvements in reliability and security.