Now available to civil as well as military users, controlled reception pattern antennas boost resiliency for many GNSS applications.

Since 2015, controlled reception pattern antennas (CRPAs) have emerged in the civil applications market, where the need to combat GNSS signal jamming and spoofing has grown exponentially. Previously available only to authorized military users, these powerful antennas can significantly increase resiliency for many GNSS applications, from aviation to critical infrastructure to autonomous vehicles and drones to the monitoring of heavy-duty freight shipments. 

As a particular example, consider Galileo Public Regulated Service (PRS) receivers aboard navigation systems and platforms, equipped with GPS/Galileo-compatible CRPAs. Galileo Open Service (OS) receivers in sensitive applications such as drone flights and other autonomous navigation will likely also require such interference-rejecting capability.

While new CRPA prototypes and products proliferate over the coming years, the need for advanced simulation and testing capabilities for them will steadily increase, as well as the need for evolving test methodologies. The capability to manufacture CRPAs with low-cost hardware serves to further extend their potential applications, and to complicate scenarios for which they must be tested.

To ensure reliability and integrity, CRPA systems must be thoroughly tested in a range of scenarios, using varied testing methodologies. This presents significant challenges for both the testers and the product design, engineering and integration teams. Specialist knowledge and expertise are de rigueur.

For 22 years, M3 Systems has specialized in tests and measurements, innovating with precision and passion under rigorous requirements and in challenging test scenarios. M3 Systems offers advanced solutions in laboratory-controlled testing, simulation solutions, record and playback solutions and hybrid approaches.

Why CRPA? And How?

GNSS is very vulnerable to signal interference, both intentional—such as jamming, spoofing and meaconing—and unintentional. As jamming, spoofing, and meaconing techniques have grown more and more sophisticated and far more frequent, the existing GNSS interference rejection techniques have not proved sufficient to combat them.

CRPA is a state-of-the-art solution to interference rejection, intentional as well as unintentional. The multiple antenna elements that can be controlled individually (thus the term “smart antenna”) detect the presence of interference signals and adjust the elements’ reception patterns to minimize or eliminate RFI impact. 

CRPA options include antenna null forming in the direction of the antenna source, beam steering to direct gain towards genuine signals, minimum variance distortionless response and more. Naturally, CRPA simulation and testing must exhibit a similar state-of-the-art quality in order to offer a means towards CRPA implementation in the GNSS product chain.

All new systems incorporating these new-capability antennas must be thoroughly vetted, at every stage of product development, against revamped vulnerabilities. M3 Systems has a thorough background in efficient radio-frequency interference (RFI) simulation testing, including modeling the interfering transmitters and simulating moving interfering transmitters.

In addition, such testing should take into account the potential impacts of new signal authentication schemes, already underway with Galileo’s Open Service Navigation Message Authentication (OSNMA) and future testing of GPS modernization proposal to take place aboard the NTS-3 satellite, namely Chips Message Robust Authentication (CHIMERA).

Note that CRPAs constitute one methodology of multi-element antenna techniques against jamming and spoofing. Flexible and configurable testing, such as M3 Systems is highly adept in, will be needed to explore such additional technologies.

CRPAs’ Exciting Future

Developing GNSS receiver architectures will present more—and more advanced—integration between the antenna and the receiver. As this trend develops, CRPA antennas will gain further capability and intelligence to meet the demands of more exacting applications. Such high-requirement applications will drive the need for PNT component and sensor integration, as other positioning technologies augment and complement the GNSS receiver and antenna. This gives rise to further RFI challenges within the multi-varied PNT component device.

Overall, this forms a compelling case for miniaturization of all components, and the need for system-on -chip (SoC) innovations.

Needless to say, every innovation, every addition of a new component or integration structure must be thoroughly tested, in simulation first, and then in the field. Proper simulation is essential to identify weaknesses, prevent system failure and ensure continuous operations. As jamming and spoofing methods increase in sophistication, so too must test of countermeasures such as CRPAs. The factors involved in advanced counter-interference techniques are very complex, and their testing present many challenges. 

Thoroughgoing experience as well as highly developed, highly proven technology is a must. M3 Systems supplies them all.

As CRPA techniques develop and advance, they will explore three possible implementation forms, depending on the receiver layer, advantages and limitations:

• RF layer CRPA-GNSS: To the best of our knowledge, all current CRPA-GNSS product consist of this format (Figure 3).

• Pre-correlation layer CRPA-GNSS (Figure 4).

• Post-correlation layer CRPA-GNSS (Figure 5).

The next-generation CRPA-GNSS will be based on innovative mixed implementations of these techniques, and simulation of these new products must continually evolve to match their sophistication.

CRPA Simulation and Testing

M3 Systems employs a systemic layer approach in GNSS simulation. This makes it possible to generate observables at intermediate levels, such as raw data and IQ baseband. As a result, the test performance verification is simplified.

There are several important key performance indicators (KPIs) and topics to assess before undertaking CRPA testing:

• Characterization of individual antenna elements

• Phase alignment: it is required that test bench is phase aligned (below 5ps at minimum & phase-sharing architecture) to keep the CRPA unit under test able to reject undesired signal

• Anechoic chamber limitations: The time of validity is limited to tens of minutes. CRPA simulation in a controlled environment offers many advantages compared to over-the-air (OTA) testing with an anechoic chamber. With OTA testing, the scenario duration, time and date are limited by the fixed positioning of transmit antennas; angular fidelity is acceptable for a few tens of minutes.

The key parameters for a testing campaign with CRPAs include:

• the power and carrier-phase calibration of the system

• the number of frequencies and constellations to be tested 

• signal fidelity and spectrum purity

• sensor fusion.

The main challenges for CRPA testing encompass:

• Which KPI should predominate in simulation, according to the product and its intended application. For example, extent of coverage or signal fidelity? A system layer approach such as M3’s is mandatory here.

• Record and playback testing must consider the power dynamic and how to address it, as the jammer and the receiver will likely have a very high power difference. Multiple RF stages are necessary. 

• In hybrid testing, with the injection of synthesized signal phenomena on 4 antenna channels instead of one (for a 4-antenna CRPA), the necessary phase coherence and synchronization must be accounted for.

Conclusion

Because CRPAs can adapt dynamically to interfering signals, they offer a very effective anti-jamming and anti-spoofing solution. They form a crucial element for the future of GNSS. CRPA in the civil market is both achievable and practical. Product research and development is ongoing and very exciting, but it requires many means of experienced, qualified, multi-element and multi-method testing and simulation to realize CRPA’s potential improved capabilities. M3 Systems has the specialized expertise and knowledge to fill this role.

For more information visit, M3 Systems.

The post CRPA for GNSS: Benefits, Challenges and Testing appeared first on Inside GNSS - Global Navigation Satellite Systems Engineering, Policy, and Design.

CAST has built a long list of faithful clients, working with them to develop custom software and hardware and producing many unique tailored innovations.

This year, CAST Navigation celebrates its 40th anniversary of delivering powerful GNSS/inertial simulation systems to a range of clients across military and commercial PNT sectors, helping them make the decisions necessary to refine and improve their products and processes. This has been the company’s sole focus throughout its history: developing customized perfection in the single area of GNSS simulation. 

The company started in 1981 and from the outset employed engineers with experience in both inertial technology and GPS receiver design. Founders Dick Gibson, a legendary GPS pioneer and educator, George Gutheim and John Clark, Sr. quickly established a reputation for thought leadership and powerfully accurate testing tools. From an early focus on aviation, they expanded to produce embedded GPS/INS (EGI) integration tools, diagnostic tools and support equipment, controlled reception pattern antenna (CRPA) testing and more.

Over the years, CAST has built a long list of faithful clients, working with them to develop custom software and hardware. CAST’s strong R&D initiatives have produced many unique software and hardware innovations, from helping military clients land UAVs on moving aircraft carriers to testing GNSS systems in commercial land vehicles and aircraft.

MODULAR DESIGN

CAST has regularly delivered some very large systems, custom tailored for its clients’ product development labs. They are modularly designed so that clients can pick and choose the options they need, “similar to a Lego set,” says VP John Clark, Jr. with a smile.

An early milestone in company history was the development of a software simulation program for Kalman filter designers, called NavSim. The Australian Air Force approached CAST, looking for a way to simulate in their labs. CAST modified NavSim for their needs; that was the very first CAST simulator. To do so, the company adapted a satellite chassis simulator by Stanford Telecommunications, a pioneering company in the industry, founded by people largely responsible for the design of GPS signals. 

“We came to an agreement,” recalls Clark. “We used their generator. They built an interface into it so CAST could send digital data and produce RF signals. It was two racks, three feet high, a dynamic GPS simulator that did 5 channels.”

“CAST wrote all the software to generate the pseudoranges and supplied it to the 7200’s hardware. It produced L-band signals from that. CAST was the first commercial company to produce that software.”

After the company integrated the STeL generators for a period of time, it contracted with Rockwell, who had built a signal generator card that could be put into a PC and generated a complete 10-satellite GPS set of data. “We had them custom-design a control board, to enable CAST to properly drive their signal generator card and control the power levels of each satellite.”

Lou Pelosi, another VP with 25 years of experience at CAST, picks up the story from there. “We used those cards for a number of years. In 2005 we developed our own signal gen card that is FPGA-based. It’s CAST-designed, CAST-manufactured.”

“One of the lessons we learned with Rockwell cards: we can do more than one trajectory at a time, we can simulate trajectories of more than one vehicle. We’ve also learned how to do multiple vehicle IMUs based on those same lessons. It taught us how to build our software and our systems in a modular fashion. We can now drive multiple vehicles with phased array multiple element antennas. We could never do this today without learning those lessons from the past.”

“We now manufacture our own inertial interface card. We can drive almost everybody’s IMU using our own card.”

JAMMING AND SPOOFING

The company naturally focuses on both GPS jamming and spoofing, meeting the urgent needs of its customers. Its systems can drive multiple IMUs and EGIs. “We also have the capability to connect multiple systems together, to drive a whole fleet of navigators. That’s a pretty big deal.”

CAST systems are very stable, requiring minimum calibration and minimum warmup time. 

Increasingly, customers need specific jamming waveforms to fulfill their needs. CAST has the ability to deliver larger systems to drive multiple navigators, with each navigator using a CRPA antenna, and each output from each of its EGI systems coherent to one another.

THE FUTURE

The trend they see and are actively pioneering is the production of bigger systems, capable of driving multiple IMUs and multiple CRPAs simultaneously. The custom, modular approach is necessary for customer requirements in highly complicated laboratories. CAST hands over a system that is both highly developed yet able to be further customized within the customer’s own lab.

“There’s a lot of stuff they don’t tell us. These guys are developing next-gen fighters and bombers, both manned and unmanned,” says Clark.

“We can customize our interfaces to whatever anybody decides to build into their box,” he concludes. “We build our own IMU and GNSS cards, we have full control, they’re all built on FPGA.”

“We’ve learned to be flexible under different conditions and difficult conditions.”

The post Four Decades of Close, Customized Client Collaboration: The CAST Tradition appeared first on Inside GNSS - Global Navigation Satellite Systems Engineering, Policy, and Design.

Selecting GNSS and Inertial for Intelligence, Surveillance and Reconnaissance

Inertial technology coupled with GPS/GNSS plays a critical role in the fast-growing field of intelligence, surveillance and reconnaissance (ISR) applications. This article covers the key factors to consider in selecting the best GPS/inertial integration for your particular ISR job.

With any tool, you need the right one for the job, and inertial technology comes in many forms, with many different capabilities. If optimally selected, your GNSS-aided Inertial Navigation System (GNSS/INS) will form the linchpin in positioning your ISR sensors for best results. Keys to consider are differences in errors that will inevitably be encountered: determining, relative to the desired level of performance, which errors are tolerable and which are not. This is called the error budget and is the first thing to review with knowledgable equipment suppliers.

ISR technology finds increasing application in military, law enforcement, search and rescue, border patrol, disaster response and many other fields. In most cases, an airborne camera gathers data key to the operation, to locate, map and geo-reference points of interest. 

Typically, an electro-optical and/or infrared (EO/IR) camera is mounted on a gimbal, a pivoted support that rotates the camera around a single or a double axis. Gimballed systems are common on UAVs, aircraft and marine vessels. They enable pointing the sensor payload toward objects of interest, independent of the platform’s attitude or movement. That independent motion also allows for stabilization (or regulation), decoupling the high-frequency motion and vibrations of the platform or vehicle from those of the payload.

For a variety of reasons, gimballed payloads incorporate their own position and attitude (heading/yaw, pitch and roll) sensing in most applications, rather than relying on the platform’s GNSS/INS. The aircraft’s navigation system may not be accurate enough for geo-referencing objects seen from as far as 10,000 meters. In some cases, the aircraft’s navigation system may be mission-critical and thus unavailable to the ISR sensors. Finally, flexing of the aircraft introduces errors since the platform’s INS is typically mounted near the center of mass of the vehicle and not co-located with the gimbal.

POSITIONING THE SENSORS 

For most gimbal-pointing applications including geo-referencing, the gimbal control system requires position data, which precludes a pure Inertial Measurement Unit (IMU) or Attitude Heading Reference System (AHRS) solution; it requires a GNSS-aided option. Even for applications where positioning is not required, the high-accuracy pointing requirements during dynamic motion mean that the degraded performance an AHRS suffers during such dynamics is unacceptable; again, a GNSS-aided solution is needed.

A GNSS/INS system is composed of inertial sensors (which come in a variety of grades), a high-sensitivity GNSS receiver, and Kalman filtering and algorithms. Measurements from the GNSS receiver are coupled with the inertial measurements to provide position, velocity, and attitude estimates of higher accuracy and better dynamic performance than a standalone GNSS, INS or IMU/AHRS can provide. 

A GNSS/INS system typically includes a 3-axis gyroscope, a 3-axis accelerometer, a GNSS receiver, and sometimes a 3-axis magnetometer to determine a navigation solution. Each of these sensors contribute different measurements—and different potential errors—to the GNSS/INS system.

Combining GNSS and INS enables them to complement each other, overcoming some of the limitations they each face as standalone systems. A few issues remain to be addressed in an integrated GNSS/INS system. 

This article covers how to select the right GNSS/inertial unit combination for a specific application, and how to integrate that into the full range of pointing, sensor, and geo-referencing systems.

DO I NEED A GNSS COMPASS?

A GNSS/INS provides accurate data under dynamic conditions, generally for platforms moving at 5 meters/second (roughly 11 mph) or faster. A single-antenna GNSS component is generally reliable for this type of application, being able to track heading through a process called dynamic alignment. This correlates accelerometer measurements with GNSS measurements to track heading in the presence of horizontal acceleration. This is not the same as course-over-ground of a GNSS system, but is rather the true heading of the system.

Since it relies on dynamic alignment, a single-antenna GNSS/INS will lose its ability to accurately track heading in applications where the platform is stationary or moving slower than 5 meters/second, such as helicopters, rotor UAVs, or marine vessels. Even large aircraft that endure long periods of straight and level flight may find that their attitude solution wanders during such periods of low dynamics. In these scenarios, a GNSS/INS with integrated GNSS-Compass is preferred. 

A GNSS-Compass is a system composed of two GPS receivers connected to two external GNSS antennas some fixed distance apart. Differential GNSS calculations are used to determine the relative position of the two GNSS antennas to millimeter-level accuracy to measure the heading of the system. The heading accuracy of this GNSS-Compass is inversely proportional to the separation distance between the antennas.

VectorNav’s VN-300 and VN-310 GNSS/INS products are examples of dual antenna GNSS/INS systems. Advanced GNSS/INS systems with integrated GNSS-Compasses such as the VN-300 and VN-310 can automatically transition between the dual-antenna heading solution and dynamic alignment to output the most accurate heading, depending on real-time motion of the systems.

STANDARD GNSS VS RTK AND INDUSTRIAL VS TACTICAL GRADE INS

Inertial sensors come in different grades, depending on their gyroscope-induced error budgets. ISR applications typically require either industrial or tactical grade IMUs. There can be an order of magnitude difference between them. 

RTK involves use of a base station on the ground in addition to the GPS/GNSS receiver onboard the aircraft, and yields a higher degree of positioning accuracy. RTK has its limitations, however: extra size, weight and power, cost, additional hardware with base station, range to base station, and so on. 

Figures 1 and 2 show the image overlay uncertainties contributions from the GNSS/INS, depending on industrial-grade vs tactical-grade inertial performance, and standard L1-only GNSS vs external RTK GNSS positioning. 

Since heading accuracy is much worse than pitch/roll accuracy, Figures 1 and 2 also show the uncertainties based on whether the gimballed camera is generally downward-pointing (yaw errors negligible) or horizontal (yaw errors dominate). At short ranges (<100m), the GNSS accuracy dominates the system accuracy, whereas at long ranges (>1000m) the INS performance is determinative.

ISR SYSTEM ERROR BUDGET

The estimated pointing solution of the gimbal (˜PG) can be found by taking the offset of the feature in the camera (xˆ, yˆ, fˆ) and rotating it to a North-East-Down (NED) coordinate frame using an estimated coordinate frame transformation matrix (C) and multiplying it by the slant range (L) and adding it to the position of the aircraft (˜PA) as shown in Equation 1.

Each term in the equations is subject to different errors from different sensors. The error budget determines the contribution of all data sources—each sensor has an inherent margin of accuracy—that affect the acquired data quality, to check if the degree of uncertainty in the measurement solution meets the minimum job specifications. 

An error budget quantifies all of the sources of error in a system, and accumulates them into a total system performance. By understanding how certain sources of error affect the ISR system, as well as the magnitude of those errors, total system performance can be improved through smarter design choices and error mitigation techniques.

It is useful to split errors into NED coordinate frame errors and errors orthogonal to the slant range. Direct errors in position are usually referenced in an NED coordinate frame, while errors related to attitude can be represented as perpendicular to the slant range.

For all errors related to attitude (∈ [°]), the error in position (σ¯ [m]) can be calculated as a function of some distance (d) as shown in Equation 2. This can be used to find the error perpendicular to the slant range.

GNSS/INS PERFORMANCE IMPACT

A GNSS/INS introduces a measure of error into the positioning and heading solutions for the sensor suite and its produced data. The error budget from GNSS/INS uncertainty can be split into GNSS/INS position error and GNSS/INS attitude errors.

Position errors largely reflect errors in the GNSS solution. These are caused by satellite orbit data inaccuracies, satellite clocks, receiver noise, ionospheric delay, tropospheric delay, and multipath, or reflected signals. GNSS horizontal and vertical position errors translate directly to those same errors in the pointing solution.

Meanwhile, attitude errors are largely determined by the INS. Inertial sensors are subject to several common error sources such as bias, noise density, scale factor, misalignments, temperature dependencies, and gyro g-sensitivity.

The contribution to the error budget from the attitude uncertainty can be calculated by finding the error perpendicular to the slant range. The exact translation from roll/pitch/yaw to position errors in the slant range frame (dependent on gimbal look-angle, α) can be found using Equation 2 and the values in Table 1.

Multiple sources of INS error must be accounted for when determining how much error is acceptable for an accurate navigation solution (see Figure 3).Although not described in detail here, these need to be accounted for by professional guidance in manufacturing and installing a GNSS/INS on your platform. VectorNav can provide such guidance. 

OTHER ERROR SOURCES

In addition to the error contributions from the GNSS/INS, a system designer or integrator must consider a variety of other terms in the error budget, including misalignments, timing errors, and camera distortions.

Misalignments between the navigation system and the imager electronics sometimes exceed the attitude errors of the GNSS/INS system itself, and their effect can be characterized similarly attitude errors from GNSS/INS, using Equation 2 and Table 1.

Time synchronization between the EO/IR system and the GNSS/INS is also critical. Any errors in timing will be multiplied by the linear and angular velocities of the gimbal to create additional position and attitude errors.

To compensate for misalignment and timing errors, the system should be calibrated on a part-by-part basis to determine the fixed misalignment offset by comparing the measured values of the system to some independent source of truth. We recommend that this calibration take place in a benchtop testing environment during manufacturing to prevent the introduction of other error sources during the calibration.

And of course, the EO/IR system itself will have errors that must be accounted for in the error budget, ranging from intrinsic parameters that can be calibrated, like lens distortions and focal length, to limitations on accuracy due to the resolution of the imager and the performance of the image-processing algorithms.

EXAMPLE

As an example of an ISR application we will look at the case where a surveying camera is mounted to a plane. Once an image is taken, a point is georeferenced using a flat and level digital elevation model (DEM) combined with gimbal attitude and position. Often, ISR applications call for mapping a point or a feature on an image to a DEM tied to a global coordinate frame (see Figure 4). This can be done by iteratively trying to find a slant range (L) such that the pointing solution aligns with the DEM.

Because we know that our georeferenced point lies on a 2D DEM the error only exists in two dimensions. In this case a plane will be traveling at 50m/s due North at a height of 1000m. Since the error budget of many of the error sources on the gimbal depend on slant range, the example will be assessed with the gimbal pointed 60 degrees directly in front of the vehicle with a theoretical slant range of 2000m. For comparing error the same configuration, GNSS/INS system, and Camera should all be used. 

In a flat DEM the error budget from horizontal GNSS/INS uncertainty directly comes from the GNSS horizontal position uncertainty in an NED coordinate frame. For standard GNSS this gives an error of 2m in the North and East axis. The contribution from the GNSS altitude uncertainty gets projected onto the DEM as a function of the gimbal tilt angle. The position error contribution can be reduced to centimeter level by using RTK. 

Figure 5 shows the overall resulting error budget, using all these calculations, for different combinations of two types of inertial sensors, industrial and tactical, and two types of GPS/GNSS technology, with and without real-time kinematic (RTK).

CONCLUSION

We have outlined the critical factors to consider in selecting an integration of inertial and GNSS equipment for your ISR application. Clearly, the calculations are complex and the results call for careful interpretation! And, every application has different requirements; aerial surveys conducted at 100 meters altitude by a small UAV rotorcraft call for a different solution than those done from 10,000 feet by a piloted airplane. VectorNav professionals can guide you through the selection and equally important installation process for your specific job. Go to www.vectornav.com to begin the process.

The post VectorNav: Getting It Right for Gimballed ISR appeared first on Inside GNSS - Global Navigation Satellite Systems Engineering, Policy, and Design.

The ability of a simulator to drive IMU interfaces correctly and precisely is vital to understanding, testing and validating GNSS/INS performance in high-end systems.

The development, refinement and evaluation of embedded GNSS/in- ertial navigation systems (EGIs) is a complex undertaking. GNSS/INS synchro- nization requires a precisely coordinated simulation of numerous navigation signals with a tightly coupled or ultra-tightly cou- pled GNSS/INS navigation system.

The ability of a simulator to drive IMU interfaces is often the differentiator to ef- fectively understanding, testing and vali- dating GNSS/INS performance in high-end systems.

THE CROQUET CHALLENGE

The quality of a strapdown IMU is defined by the accuracy of its gyroscopes and accelerometers as well as the signal pro- cessing pipeline and the minimization of errors such as bias errors, velocity errors, platform tilt, and errors induced by limited inertial sensor bandwidth.

While GNSS systems determine position from point to point, simulating an IMU for proper performance is analogous to hitting a croquet ball through wickets—it’s a steady and calculated bump along a path. John Clark, Vice President of Engineering at CAST Navigation, explains, “Hit too hard and you miss a wicket and head off in the wrong direction with no easy way to correct. To stimulate an IMU, you give it a starting location and you bump along with stimulus. You need to model the IMU correctly with considerations for GNSS coherence and for values such as coning, sculling and other error sources such as bias and individual sensor noise.”

Military and government labs have very strict requirements for dynamic ground testing of new navigation systems prior to flight testing. RF simulators can help recreate the conditions of a real flight. “With high-fidelity sensor error models, a simulator can include the addition of errors into the inertial stimulus thus allowing for a better overall system simulation allowing for realistic system performance measurements,” he adds.

REPLICATING REALISM

CAST Navigation’s INS Systems simulate dynamics for both the GNSS and INS por- tions of EGI and GNSS/INS systems in a coordinated and coherent way so that the GNSS and INS navigation solutions do not diverge and the solution precisely dupli- cates real world conditions (Figure 1).

While some simulators construct an INS error model and add its output to the true user state, CAST’s approach more closely imitates the operation of an INS. Essentially, the CAST INS Systems drive inertial measurements. Clark explains, “It electrically disconnects or puts the gyros and accelerometer to sleep and sends delta V and delta θ to a Kalman filter in order to test system dynamics. In this condition, the EGI can output raw IMU measure- ments to a receiver under test, which allows users to test GNSS receivers that are tightly integrated with a strapdown IMU.”

Initially, the CAST INS Systems simulate an ideal IMU by computing error-free inertial measurements using a complete whole-state model of the inertial system. Then the simulator adds errors to the measurements using detailed sensor error models and simulates the output of the IMU.

As shown in Figure 2, the user motion profile is defined by the MSF (Maneuver Segments File). The UMG (User Motion Generator) expands the maneuvers in the file into a trajectory (defined by the user state at various times). The IMU truth model uses the user state to compute error-free values of specific force and al- titude. The Sensor Data Generator then uses the gyro error model to compute error-free Delta θ measurements from changes in altitude. The accelerometer error model adds the error measurements to the ideal specific force measurements. Delta θB and Delta VB (relating to the ve- hicle body frame) are properly formatted, and output through the appropriate interface with the correct timing.

“The CAST INS Systems read the file, performs the necessary formatting, and outputs the measurements through the appropriate interface with the correct timing,” confirms Clark.

RELIABLE PERFORMANCE

As a leader in mixed-signal optics and Systron Donner Inertial MEMS-based products for aerospace and defense systems, EMCORE Corp. relies on GNSS/INS simulators for hardware-in-the-loop testing to verify the expected performance of algorithms.

Andy Williams, senior field applications engineer with EMCORE explains, “Few things can be as rewarding as a successful field test or as embarrassing and detrimental as a failed field test. Vetting the system as thoroughly as is reasonably possible is essential. Hardware-in-the-loop testing is a critical part of that vetting process.”

A proper evaluation of an GNSS/IMU device requires the IMU behavior to be accurately replicated and synchronized to the GNSS radio frequency (rf) for a given flight path, the CAST simulator provides error free IMU data to our GNSS/IMU data to our GNSS/IMU system,” he continues. “In real time, the GNSS/IMU system perturbs the data from the CAST simulator with data from the internal gyroscopes and accelerometers to replicate sensor imperfections. Simultaneously, the GNSS/IMU system receives from the CAST simulator GNSS rf that is synchronized to the error free IMU data. The GNSS receiver and navigation algorithms are tested over a variety of situations with relative ease and minimal cost.”

TACTICALLY SOUND THROUGH SIMULATION

In a recent study, EMCORE engineers sought to validate the velocity and altitude limits of a new GNSS receiver along with the algorithm performance in a tactical grade SDN500 system.

The test requires altitude over 24,000 meters and velocities over 600 m/s. Only a few aircraft in the world have such capabilities. While a test flight on an SR-71 Blackbird might be appealing to many, it isn’t practical. On the other hand, simulating the test flight with the CAST system is straightforward and cost-effective. The CAST GNSS/INS simulation generates GNSS and IMU signals corresponding to a user-defined flight trajectory.

The SDN500 receives the dynamic inertial data from the CAST simulator for the flight trajectory and combines that data with the imperfect inertial data from the IMU sensors. Simultaneously, the GNSS receiver in the SDN500 receives the synchronized GNSS rf from the CAST system. The SDN500 navigation processor uses the combined inertial data and the GNSS receiver output in the navigation algorithm to provide a navigation solution to the user.

The hardware-in-the-loop use of inertial data perturbed in real-time by the actual IMU sensors and the real-time data from the embedded GNSS receiver provides a more accurate representation of the SDN500 system performance than would otherwise be achieved by an equivalent pure software simulation.

The test began with a stationary period on the ground while the SDN500 initializes and transitions into air navigation mode. After the initial hold period, the flight trajectory entered a series of maneuvers that provided observability for various parameters with corresponding changes in the calculated figures of merit. The flight scenario then continued through a series of speed and altitude changes. 

The dynamic constraints tested include:

• Max altitude of 24,000 meters

• Max velocity of 600 m/s

• COCOM combined altitude (18,000 m) and velocity (1000 knots/514 m/s) limits

A TRUE REFERENCE

A sampling of the test results for the GNSS/INS simulation is shown in Figure 3, which begins with the ‘flight’ on the ground and climbing to an altitude of 25,000 meters (above the absolute max altitude defined by manufacturer=24,000 meters) and then dropping to an altitude of about 15,000 meters (graph 3b between 2.1-2.2).

The top graph (3a) plots the system status flags pertaining to GNSS availability to the SDN500. The second graph (3b) tracks the altitude throughout flight while the third graph (3c) shows the magnitude of velocity of the solution compared to truth reference system (solid orange line).

Throughout the test, the CAST simulation continues to test COCOM restrictions, initially with the sequence of altitude greater than 18,000 meters and then for a velocity exceeding 514 m/s. Notice both altitude and velocity must violate the respective limits before the GNSS receiver fails in order to provide the 1PPS and position/velocity solution. Next, the sequence is reversed with the velocity exceeding the COCOM limits first, followed by altitude. Critical to the test results is that the blue 3501 msg data (SDN500) and orange truth (CAST trajectory input) lines overlapped throughout even when there was no valid solution from the GNSS receiver.

Williams explains, “During the times when there was no valid solution from the GNSS receiver the algorithm maintained an accurate solution using only the data from the IMU. In addition, there was no algorithm instability or discontinuity when the GNSS receiver resumed, providing a solution to the algorithm. Throughout this entire profile, even when GNSS signal is lost, the SDN500 maintains an accurate navigation solution. This test is not possible without the synchronized GNSS rf and trajectory matching IMU data provided by the CAST system.”

Bottom line, the GNSS receiver and navigation algorithm was confirmed to operate as expected throughout the operation for all three of the dynamic constraint scenarios.

MANAGING INTERFERENCE 

The CAST INS Systems simulator has the capability to interface with GNSS/INS navigation system under test in such a way as to record the navigation messages from the GNSS receiver that report jamming relevant values. Typically these messages include J/S ratios, satellite azimuth and elevation, and sometimes even individual satellite pseudoranges. By making these recordings available to users, the test report creation process is greatly simplified.

The CAST INS system can handle up to eight independently controlled jamming signals. Each generator can be assigned a speed, altitude, and trajectory profile. The jamming modes include Continuous Wave, Pulsed Continuous Wave, Swept Continuous Wave, FM Noise, Pulsed FM Noise, and Wideband Noise. 

Properly simulating modern integrated navigation systems requires the ability to provide high rate inertial data with exceptional fidelity. CAST INS systems are equipped to test a wide variety of vehicle types including airborne, terrestrial, aquatic (including the sea-state) and space-based orbital vehicles for military, industrial, and commercial laboratory clients

The post A True Reference: Theory Meets Reality in Synchronized Simulation Environments appeared first on Inside GNSS - Global Navigation Satellite Systems Engineering, Policy, and Design.

The challenge for chip designers at Physical Logic has been to convince inertial measurement unit (IMU) and UAV manufacturers that the company’s revolutionary approach to micro-electromechanical systems (MEMS) machining can vault low cost, weight and power (C-SWaP) accelerometers several tiers above in the accuracy hierarchy. In short, MEMS can replace the traditional mechanical or solid-state accelerometer in all applications, achieving tactical and even navigation-grade inertial performance.

Key features of the innovative control design give system-wide low noise, high linearity, virtually zero vibration rectification error (VRE), robust stability and enhanced sensitivity.

“Almost all of these [traditional mechanical accelerometers] can easily be replaced with a Physical Logic closed-loop,” said Aviram Feingold, the company CEO. “If you wish to compare the systems, you will see. These are mainly for navigation, where an achievement of 70g enables us to get into maybe 90% of the end-user applications. UAVs, helicopters, missiles, marine applications, and now there are wireless applications where 30 g is not enough, they need 50g or 70g.”

Two variants of an innovative accelerometer design, Open-Loop and Closed-Loop, are based on a unique in-plane bulk micromachining process. This is performed on silicon-on-insulator (SOI) wafers, assembled with application-specific integrated circuit (ASIC) with an analog front end and a temperature sensor in a custom-tailored, hermetically sealed and compact LCC20 package.

The unique in-plane geometry of the sensor’s MEMS element enables it to use a highly symmetrical comb structure based on a full capacitive bridge topology. The in-plane concept enables excellent capacitance linearity over a large proof mass displacement. It also produces extremely low thermo-mechanical noise levels, achieved by sealing at atmospheric conditions without using less-reliable vacuum packaging.

Open Loop

Physical Logic’s Open-Loop accelerometer series fills the gap between current MEMS accelerometers and traditional, much heavier and more expensive mechanical accelerometers. The Open-Loop product line, offering a variety of input sensing ranges from 2g to 70g, finds applications in tilt sensing, seismic sensing, resource exploration, vibration sensing, North-finding and inertial navigation.

In-plane micromachining takes advantage of the normal MEMS benefits such as low-cost manufacturing and easy integration with microelectronics. However, it avoids the downsides of high-impedance readout for capacitive sensing. This readout node is highly susceptible to parasitic leakage current and electromagnetic interference. In-plane micromachining uses a full-bridge capacitive sensing configuration for parasitic rejection and improved noise performance.

It also enables improved linearity. With equally distributed sensing capacitors on the area of the MEMS die, the process further leverages its symmetric mechanical geometry to offer high stability performance. The simplicity of the design means that only minor architecture changes are required to achieve various sensing ranges.

Physical Logic’s Open-Loop MEMS accelerometers fulfill the requirements of tactical-grade inertial performance (see Table 1, Open Loop 2000 Series Key Parameters).

Closed Loop

To overcome inherent non-linearity and further improve bias stability performance, necessary to achieve navigation-grade performance, a further innovation was necessary. Thus, Physical Logic developed the Closed-Loop series, similarly in-plane micromachined, but with a rebalance force applied on the proof mass to offset the inertial force caused by external acceleration. The closed-loop design compensates for the acceleration, minimizing the proof mass displacement

Thus equipped, the Closed-Loop series goes even further, replacing tactical and navigation-grade mechanical accelerometers. With qualification of its sensing range up to 70g, it is ideally suited for high-end navigation applications, such as UAVs, other autonomous vehicles, aviation and space. It offers improved performance in the form of scale factor linearity, bias stability and vibration rectification.

For example, the MAXL-CL-3030, with 30g dynamic range and 20-bit resolution, is a fully integrated sensor enclosed in a specially designed and manufactured LCC44 package. See Table 2, Closed Loop 3000 Series Key Parameters.

Recently, Physical Logic announced the completion of qualification and production readiness of the MAXL-CL-3050, a new 50g sensing range sensor based on MAXL-CL-3000 family. As with the 15g and 30g configurations, the performance of MAXL-CL-3050 allows it to compete with traditional mechanical accelerometers in the most demanding navigation-related applications. Yet higher levels of performance are reached with the new 70g sensor.

For both open-loop and closed-loop product types, Physical Logic’s robust in-plane bulk-micromachining process assures high yield and high reliability, delivering improved performance while preserving the MEMS advantages of low C-SWaP.

A Unique Approach

Several years ago, Physical Logic undertook a dual research and development project on capacitive MEMS accelerometers, with the aim of achieving inertial tactical-grade performance for open-loop accelerometers and inertial navigation-grade performance with closed loop accelerometers. The company scientists came up with a unique design that affords the same MEMS fabrication process flow for both product lines. Further, robust process ensures high yield and high reliability.

The MEMS manufacturing process has normally followed an out-of-plane design. The resulting products require vacuum packaging, since damping of large parallel plate electrodes is very severe in the out-of-plane configuration. Vacuum sealing always has a reliability problem. The in-plane process can take advantage of atmospheric pressure sealing for improved reliability and simplicity.

Out-of-plane sensors using the gap-changing principle sacrificed linearity performance. In-plane machining does away with this, due to it capacitance area-changing architecture. And, as previously stated, full-bridge capacitance sensing architecture provides parasitic rejection and improved noise performance, both of which are very difficult to achieve with out-of- plane devices.

Finally, in-plane machining results in improved performance during vibration, an extremely important performance metric for UAVs, particularly for delivery drones.

“The main advantage is almost zero VRE,” said Aviram Feingold, Physical Logic’s chief executive officer. “The reality here is that the VRE is a magnitude better in a closed-loop accelerometer, 10 times better than the open-loop accelerometer.

“Take, for example, the application of package delivery systems, in which safety is the most important parameter. One accident caused propeller vibration affecting the navigation of a delivery drone will have huge repercussions. So this is one of our primary end-user target applications: contacts that we have with UAV manufacturers specifically making package delivery systems.”

Figure 1 shows a scheme of in-plane bulk micromachining. Both open-loop and closed-loop structures include an array of capacitive electrode plates, mechanically coupled to a common proof mass (the rotor shown in picture A) and an array of the same electrode plates attached to the fixed silicon frame (the stator). A careful design of the supporting springs constrains the proof mass to linearly move in the plane parallel to the plane of the fixed electrode plates. The capacitance between the two arrays of plates is variable and dependent on the proof mass displacement.

Closed-Loop’s High Sensing Range

Many navigation applications require input acceleration measurements higher up to 50g or 70 g. This can be difficult to achieve without losses in other key performance parameters. This necessitates very close understanding and careful management of the overall error budget.

An open-loop accelerometer’s sensing range is usually proportional to the spring constant. Increasing the spring constant for a higher sensing range also increases the bias charge sensitivity, detracting from performance. In non-navigation applications this tradeoff can be acceptable, but navigation-grade accelerometers must deliver both high performance and high sensing range.

Closed-loop operation eliminates the tradeoff between performance and sensing range by using capacitive sensing and feedback voltage to determine the sensing range, optimizing for both outcomes. A company-developed theoretical error budget thus leads to a MEMS design achieving low bias sensitivity and superior performance over temperature and time.

“One interesting application is resource exploitation,” said Feingold, “where you need underground navigation. You have no GPS time, and long measurements underground, it’s very difficult. These companies know now that we have quantified our performance, and we start to get feedback that in terms of accuracy, we provide the good results, better than our competitors.”

Figure 2 shows open-loop and closed-loop accelerometers prior to cover sealing.

“Fundamentally, both open loop and closed loop are based on the same in-plane technology that allows for a very linear transfer function from acceleration to the voltage to the capacitance that we see,” added Lisa Koenigsberg, the company’s chief technical officer. “Because we’re using not just a gap effect, the design is optimized to get as much sensitivity as possible. With an in-plane configuration, it’s much easier to optimize. That’s really what allows the good performance in the VRE and scale factor. Because we have a lot of sensitivity, it’s very linear from the core technology.

“When we close the loop, we have basically three capacitor sections that are all mechanically attached. The acceleration now is the voltage that we apply to those actuators. That’s how we can get a high sensing range, because it’s only dependent on the voltage.

“The nice thing about the closed loop is benefiting from everything and getting all the nice performance from a low range. The high range is achieved not with a motion or displacement, but rather the voltage. So you get a high range and small biases from let’s say, a 2g accelerometer, and yet you can measure up to 50g or 70g.”

Rigorous Quality Control

Production units of each open-loop and closed-loop configuration have passed acceptance and qualification tests including temperature, vibration, and centrifuge measurements up to their designated g levels. A temperature cycle fits a baseline temperature model for bias, scale factor and misalignment behavior, followed by temperature storage cycles temperature operational cycles. Dynamic tests include vibration and shock profiles in multiple directions. Finally, the units repeat the first temperature cycle to test model repeatability.

All in-plane machined devices demonstrated excellent performance over environmental exposures, verifying the long-term accuracy of the full range of parameters. Specific statistics are available in each product data sheet. In addition, the closed-loop series achieved excellent scale factor linearity, VRE and superior stability.

“Another thing we are doing here,” said Feingold, “to convince our customers is that we are running lifetime testing to accelerate the C-Swap. We actually use accelerometers for many months, under 60 degrees. Every month we take them out, do some testing and bring them back, we test again and again and again to show very nice results.”

Conclusion

MEMS accelerometer advances over the past two decades have opened up new opportunities in many areas of motion sensing. With the explosive growth of UAVs, precision requirements for highly dynamic, high-vibration platforms have increased the downward pressure on cost, weight, size and power. The unique advantages of in-plane MEMS architecture introduced by Physical Logic mean that acceleration, vibration, shock, tilt and rotation in high-g applications can be measured with greater accuracy than previously achievable in small, light form factors. Open-loop and closed-loop MEMS accelerometers can now replace high-cost instruments in most high-end applications.

The post Achieving More with Less (Cost, Size, Weight and Power) Innovation Accelerates Accelerometers into Higher Levels of Inertial Performance appeared first on Inside GNSS - Global Navigation Satellite Systems Engineering, Policy, and Design.

L3Harris has it available for U.S. armed forces across all programs and platforms.

The post You Need M-Code. We Have It Now. appeared first on Inside GNSS - Global Navigation Satellite Systems Engineering, Policy, and Design.

Half an hour later, the father dropped the keys on the dealer’s desk. “It’s got everything we want—almost,” he said regretfully. “I’ve got to tell you, the navigation just doesn’t cut it. We were down- town, and the dashboard display showed us all over the place. We could not get reliable voice directions. She had us going forward, back, over, under, sideways. Totally confused down between the skyscrapers. I can’t put my teen drivers, my whole family, into something unreliable like that. We’ll have to look somewhere else.”

The post QuNav: Navigation That Satisfies, Navigation That Is True – Software, Low-Cost Sensors Solve Vehicle Problems appeared first on Inside GNSS - Global Navigation Satellite Systems Engineering, Policy, and Design.