How to test complex active GNSS antenna systems?
The ability to respond correctly to jamming and spoofing attacks is a key feature of GNSS receivers in critical environments. The threat of interference to sensitive GNSS signals is increasing, both from civilian sources, such as illegal privacy devices designed to disguise a vehicle's location, and from the use of radio frequency interference (RFI) deployed as a method of electronic warfare to disrupt an adversary's operations.
In the military sector, the need to protect mission-critical PNT systems has led to the development of adaptive antennas that can be retrofitted to a wide range of GPS receivers. Adaptive antennas are also increasingly being used in commercial applications such as surveying, mining and autonomous vehicles.
Testing adaptive antenna systems, especially if they are to be used in safety and liability critical contexts, is fraught with challenges, from defining test requirements, to selecting test equipment, to conducting a test using a radio frequency constellation simulator.
Adaptive antenna systems
Adaptive antenna systems come in various forms, the basic function being that the antenna maximizes the reception of true GNSS signals while attenuating the effects of interfering signals.
Attenuation can be achieved both by the physical design of the antenna and by algorithms that control the behavior of the antenna. Adaptive antenna technologies include fixed radiation or reception pattern antennas (FRPA), which are protected against signal reception from a specific direction by a mechanical barrier, and controlled radiation or reception pattern antennas (CRPA). CRPAs consist of several individual antennas that can steer their reception pattern in real time towards the real signals and away from the interfering signals. CRPA antennas are usually planar in design, with multiple antennas arranged in a circle around a central reference element. Studies have shown that a convex design offers the additional advantage that the radiation pattern can be steered both vertically and horizontally. The steering algorithms work by assigning and changing the weighting for each antenna element based on the received phase shifts of the individual signals. By changing the weighting, the radiation pattern of the array can be altered to direct nulls in the direction of the unwanted signals or gains in the direction of the true signals. The combined output of the antenna electronics is forwarded to the GNSS receiver for further processing to determine position, velocity and time (PVT).
The steering algorithms can be embedded in the GNSS receiver or controlled by a separate antenna electronics unit. The best performance is usually achieved when the algorithms are embedded in the receiver (Rx) and the system is integrated with one or more inertial measurement units (IMUs).
Challenges in testing adaptive antennas
Testing adaptive GNSS antenna systems involves some complicated tasks. The capabilities of the antenna and receiver need to be tested in realistic scenarios and ideally with the ability to repeat the same test conditions many times.
This requires an initial risk assessment to understand the range of interference scenarios the system may encounter - including possible future scenarios - and the desired behavior of the system in response to these scenarios. A test plan must incorporate these scenarios to understand their impact on system performance. The appropriate test method must be selected at each stage of product development. There are a number of options, some of which require a significant investment of time and budget.
Appropriate test equipment - hardware and software - must be selected to ensure that testing can be carried out accurately and reliably in a laboratory, chamber or open-air environment and that future testing requirements can be accommodated. Equipment must be properly installed and configured, which can require considerable expertise, especially for more complex chamber-based tests.
Tests must be set up and performed correctly, and results must be closely monitored and recorded. Extensive testing may require a degree of test automation, which also needs to be set up correctly.
Test planning
An important part of test planning is the development of test scenarios based on the threats that the system or device is likely to face in operation. A risk assessment determines which antenna and receiver functions need to be tested, with what level of rigor and in which scenarios.
Faults that may affect the device include:
Jamming
Both intentional jamming of GNSS frequencies and unintentional interference from radio transmissions in frequency bands close to GNSS frequencies are increasing. Vehicle owners are acquiring illegal jamming transmitters to override the on-board telematics. The increased military use of RF signal jamming in conflict zones is also having an impact. Interference from adjacent frequency bands is becoming more frequent as frequencies adjacent to GNSS bands are allocated to other services.
Spoofing
GNSS spoofing, the transmission of fake but deceptively real GNSS signals to the receiver, used to be a rare jamming technique but has become easier and cheaper with the advent of software-defined radios (SDR). An inexpensive spoofing device can be built from components found on the internet and downloaded open source code.
Obscuration
GNSS signals, which work on the line-of-sight principle, are blocked by objects on the ground such as tall buildings, slopes and dense foliage. A GNSS-based system is vulnerable not only when it cannot receive satellite signals, but also when it leaves an obscured area, such as a tunnel or underground parking garage. When attempting to receive a signal again, it may be subject to a spoofing attack, causing it to tune into the fake signal rather than a real one.
Multipath
In addition to the direct (line of sight) signals from the satellite, the signals can take several paths: They are reflected or diffracted by buildings or other objects in the vicinity. These signals have a slightly longer path to cover and therefore arrive at the receiver somewhat later than signals with line of sight. Without mitigation, multipath signals can cause the receiver to output an inaccurate distance measurement, resulting in an incorrect position.
Different test methods
Wired tests with simulated signals and interference
In a wired test, all relevant elements of the RF environment are transmitted directly to the antenna electronics via a coaxial cable. The signals from a single or multiple satellite constellations are generated by an RF constellation simulator (RFCS), which optionally includes multipath interference and signal masking effects as well as atmospheric interference. Disadvantage - the tests performed bypass the physical antenna so that the effects of the antenna behavior on the receiver are not evaluated.
Spirent image: Configuration of the simulation of a CRPA antenna
Live Sky" tests on an open-air site
The tests are conducted in an open-air site with live satellite signals and with real jamming and spoofing equipment (with the appropriate regulatory approvals) to jam the signals from space.
The richness and authenticity of the real environment provides a reliable indication of the real-world performance of the antenna and receiver in the presence of RF interference. However, when testing in the open air, there is a risk of collateral damage to GNSS-dependent systems in the vicinity of the test area. To mitigate this risk, the power of the jammer and spoofer often needs to be reduced. The performance of the antenna and antenna control systems can still be evaluated, but because the scale of the setup is compromised, the angle measurements may be less accurate than desirable. In addition, outdoor field trials can incur significant costs in terms of time, resources and equipment. The real-world RF environment is constantly changing, and conditions such as satellites in line of sight, multipath effects and temperature cannot be controlled, meaning that test conditions are never exactly repeatable. This leads to accuracy and reliability issues when tests are performed iteratively.
Absorption chamber with fixed antenna array
The conventional method of configuring an anechoic chamber for CRPA testing is to place a transmitting antenna (Tx) at the same azimuth and elevation as the satellite to be emulated. The antenna then transmits the signals from this satellite using a single-channel GNSS signal simulator per antenna. Interference sources such as jammers and spoofers can be placed in the chamber.
For added realism in dynamic scenarios, the Rx antenna can be mounted on a turntable/3D positioner that replicates the attitude changes of the simulated vehicle platform. For units such as handheld devices, objects such as reflectors and signal attenuators (e.g. a human dummy head) can be placed
can be placed next to the DUT (device under test) to simulate a realistic environment. Multi-path effects, obscurations and atmospheric interference can be introduced using advanced 3D environment modeling and ray tracing. The advantage is complete control over the test environment so that the same conditions can be repeated to test different antennas, antenna designs or antenna arrays. The antenna hardware is included in the test, along with the potential for different angles of arrival of the signals on the hardware.
However, the limitations of the chamber design make it difficult to accurately replicate the geometry of a particular GNSS constellation or to simulate an environment with multiple constellations
Fixed transmit antennas are not capable of mapping the movement of satellites in orbit, but rather represent a specific fixed location, time and date. Only very short test scenarios are possible (around 30 minutes), as the environment quickly becomes unrealistic. This can be used to evaluate the receiver's ability to acquire and track GNSS signals, and it can demonstrate the beamforming capabilities of the CRPA, but it is not an effective method for testing an active or responsive antenna system.
Zone chamber with simulation of GNSS satellites in orbit
A zone chamber also uses fixed antennas distributed in a regular pattern of azimuth and elevation angles. However, unlike configurations that require a fixed transmit antenna and a single-channel simulator per satellite, the zone chamber uses one transmit antenna per zone from which all signals in that zone are broadcast.
The signals from a particular satellite are broadcast in one zone before being switched off in that zone and immediately moving on to the next, emulating the pattern of satellite movement in orbit. By realistically simulating specific constellations of satellites orbiting the Earth, the zone chamber concept represents the state of the art in OTA CRPA testing.
The size of the zones can be customized according to the test requirements and the dimensions of the chamber. Spirent's experience with existing zone chambers has shown that 31 zones is the optimum configuration. A smaller number of zones at lower altitudes and a higher zone density above 10° can provide a better representation of the orbital paths.
A smaller zone size provides a better representation of the orbital paths. The reason for this is that the signal angle becomes less accurate as the orbit increases towards the zone edge and a smaller zone size shortens the distance between the drilling site (in the center of the zone) and the zone edge. However, the disadvantage of smaller zones at higher altitudes is that there are more transitions or handovers from zone to zone.
Each satellite signal is evaluated independently. Although there is a small signal interruption when signals change zones due to an unavoidable discontinuity in the carrier phase, this only affects the signal from the satellite in question. This can lead to a momentary loss of carrier lock on the affected satellite, but these events are rare and it is unlikely that more than one satellite will be affected at a time. The extent depends on the quality of the calibration. The frequency of change increases with the number of zones; for a 31 zone system a typical frequency would be every 2-3 hours for a static vehicle.
As with the fixed chamber, interference sources such as jammers and spoofers can be placed anywhere in the chamber. A rate table/3D positioner, reflectors and signal attenuators can be used to create additional realism. Realistic multipath effects, obscurations and atmospheric interferences can be designed using advanced 3D environment modeling that takes into account the arrival angle information of the signals and manipulates it accordingly.
A zoned chamber makes it possible to simulate the movement of real GNSS constellations in orbit. Individual constellations (e.g. GPS, Galileo, GLONASS) can be simulated, as well as combinations of constellations (e.g. GPS + Galileo + GLONASS). The test scenarios can run for longer without losing realism. This makes the zoned chamber suitable for validating all aspects of the CRPA system, including beamforming, null steering and SFAP/STAP processing. Azimuth and elevation of each satellite can be realistically replicated so that direction finding and antispoofing capabilities can be reliably evaluated.
The zone chamber concept was patented by Spirent in 2014 under the title "Over the air GNSS testing using multi-channel generators to create spatially dispersed signals" (USPTO patent no. 8854260).
Author: Spirent Communications
Translated from English and abridged by: Lange-Electronic GmbH, Daniela Knöferl
All images @Spirent
Original Whitepaper:
Characterizing CRPAs and other adaptive antennas
How to test CRPAs and other advanced GNSS antenna designs
Download: https://nextcloud.lange-electronic.com/s/KDAmfc4cexWHjmY
