An EMI Analysis of NXP HoverGames UAV

I. Introduction

Due to the countless flight problems, I have had with the HoverGames drone (loss of control, unexpected and unjustified crashes, even I lost a drone in a wooded area, strange flight behaviors from time to time or on long intervals of time, etc.), I decided to do an analysis of the electromagnetic disturbances generated by the various components and how they affect the GPS module. So, this study analyzes mainly L-band EMI interference generated by different components placed on an NXP HoverGames UAV, a key aspect in understanding and preventing failure in the system's GPS correct functioning.

Nowadays, many unintended frequencies leak out all the time due to both an increasing crowded radio spectrum, largely populated by the new and in full ascension 5G standard and to the more and more complex and sophisticated electronic devices, countless types of gadgets, electric cars etc.. Thence, all electronic devices are exposed to a polluted environment, with even increased and more rich electromagnetic interferences (EMI) than ever before.

In a drone, EMI commonly affects the GPS ability to receive the transmissions from the low-power multiple GPS satellites and, also, impacts the magnetic field received by a magnetometer. Accordingly, the drone's position and orientation are disturbed due to the GPS signal loss and heading instability, respectively.

The effect of different EMIs on the position and accuracy of GPS receivers has been widely reported in the literature. For example, the EMI noise, generated by the SSD (Solid State Drive) and USB 3.0 systems, adversely affects the receiving performance of GPS modules from within the portable devices [1] like notebooks or ultrabooks. As a direct result, such devices' position and navigation performance degraded. Furthermore, onboard computers are another very powerful source of EMI [2]. For example, in a CubeSat nano-satellite, the internal GPS receiver could not get a fix when the satellite was completely assembled. This problem was foreshadowed in the initial tests when the GPS receiver was unable to get a position fix in less than 10 minutes while the expected time to fix, accordingly to the GPS receiver datasheet, had to take no more than 2 minutes. In the end, as a result of the research done, the problem was linked to three different components [2]: the onboard computer, which was the source of EMI, the solar panel, which was disrupting the antenna near-field, and the GPS antenna itself.

Based on my previous experience in developing a UAV system for video monitoring of a quarantine zone, several sensitive aspects could be further taken into discussion [4]. The first one is related to the range and the object penetration (e.g., bridges, trees, buildings, etc.) done by the radio control (RC) and telemetry links. The second one, more important and more dangerous than the first one, consists in the EMI generated by the drone components (motors, onboard computers, electronic power systems, etc.). This study presents several solutions and analyses to understand better and solve these issues.

II. HoverGames UAV system

In the original configuration, the HoverGames drone has an RC link channel that uses a FlySky FS-i6S RC transmitter and an FS-IA6B receiver – a radio of a 2nd generation AFHDS (Automatic Frequency Hopping Digital System) protocol, able of 2-way limited telemetry communication. This link has two main problems: (a) the latency increases with range, and (b) it uses a 2.4 GHz radio band. A lower frequency for the RC link will ensure a more extended range and a higher penetration of the obstacles. As a direct result of this new implementation, a TBS Crossfire 868 MHz long-range RC link was used. Based on this link, the maximum range is more than 100 Km, but the realistic range is a little more than 25 Km.

The telemetry channel provides a wireless MAVLink connection between a ground control station (QGroundControl in our case) and the drone. Based on this link, inspecting different real-time drone parameters, tracking the drone on the map, or changing a mission on the fly becomes quite possible. Based on the actual setup, we have two options to implement the telemetry link. The first one is based on the TBS Crossfire RC protocol (a TBS Crossfire Diversity Nano RX being used in this sense), through which a telemetry link could be tunneled between the drone and the base station. The second option resides in using an external module connected to the FMU (flight management unit) – in our case, HolyBro HGD-TELEM433. In what follows, the selection of one of these two proposed options will be done based on the EMI generated by these systems.

The HoverGame drone uses a HolyBro Pixhawk 4 module, composed mainly of a UBLOX Neo-M8N GPS module and an IST8310 compass unit. Neo-M8N GPS module can concurrently connect to three Global Navigation Satellite Systems (GNSS) constellations from the ones supported by the module: United States' GPS/QZSS, Russia's Galileo, China's BeiDou and the European Union's GLONASS. All of the GNSS operate in L-band (i.e., 1-2 GHz). The signals emitted by GNSS are at a very low power level, which makes the signals susceptible to interference from other signals transmitted in the GNSS bands. When the interfering signal is powerful enough, it becomes impossible for the GPS receiver to detect the low-power GNSS signal. Moreover, if an object blocks the GNSS receiver's line of sight to a GPS satellite (buildings, trees, etc.), the signal from that specific satellite will not be received. In such a case, the solution is to track more than one constellation. Thus, more satellites will become available which, in turn, will provide a better chance of finding enough satellites to determine an accurate position at all. In our practical case, in all the analyses, the GPS receiver acquired a minimum of 14 satellites and a maximum of 21. Such a large number of acquired satellites has the benefit of protection against the EMI by tracking multiple frequencies and multiple constellations. For example, if the EMI is in the GPS L1 band (i.e., 1550 to 1600 MHz), the receiver can still provide position by tracking GPS L2 or L5 bands, GLONASS L2 band, or Galileo E5 band.

III. Disturbance sources for the GPS module

Inside a drone, the primary EMI is given by the electric and magnetic energy emitted from the changing voltages and currents, which are passing through different circuits. The higher the voltage and the current, the stronger the electromagnetic fields will be, and therefore a greater EMI will be generated. Each drone has its own EMI background generated mainly by the brushless motors, the ESCs (Electronic Speed Controller), the power distribution system and electronic units FMU, as well as, the existing on-board computers. Each on-board computer has its own specific electric and magnetic field (EMF), which interferes more or less with the GPS. For this reason, here I analyze seven well known single board computers (SBCs): DragonBoard 410c, NavQ, NavQ Plus, Raspberry Pi 3B+, Raspberry Pi 4B, Jetson Nano, and Jetson Xavier NX, see Table I.

Also, there are additional perturbations of the GPS receiver. For example, the signals emitted by the onboard transmitters can generate intermodulation distortion (IMD). In this mode, two signals broadcasted together can cause interference at additional new frequencies generated inside the GPS receiver due to the non-linear characteristics of the electronic circuits. Other problems are related to the harmonics of the telemetry transmitter that can fall in the GPS receiver band.

IV. UAV EMI analysis

A. Database

For each analysis, three sets of measurements were done. Each measurement took a minimum of 2 minutes and 30 seconds. After the arming, the drone was lifted to an altitude of 3 to 4 meters, where the Hold mode was activated for stable hovering. The drone was manually kept in the hovering position only if the Hold mode could not be kept autonomously due to the EMI. When the time elapsed, the UAV was manually landed, and the drone was disarmed. To prevent any problems caused by EMI (e.g., unpredictable behaviors generated by the GPS, like drone drifting or flying away), a geofence and a kill switch, which kills the motors, were configured.

Each time when the HoverGames UAV is armed, the FMU starts to record on a local SD card; thus, with each measurement around 900 parameters are written into an ulog file. The recording is stopped when the UAV is disarmed. All the GPS-related data is recorded two times in a second by the FMU. Using the PlotJuggler time series visualization tool, the ulog files were converted to CSV format files. In the following step, a Matlab program was developed to extract all relevant data and to compute all the parameters used in the EMI analysis. The statistical parameters (i.e., mean and standard deviation) were computed for each set of measurements out of the all 3 series. In the end, a new parameter of the same type was computed based on these three values – see Tables II, III, IV and V.

B. EMI measurement

The measurement of the EMI is not an easy task. Different software programs can also model EMI generated by every system component, but these software packages are costly. So, in my case, I decided to use the GPS noise and jamming parameters extracted from the ulog database. The jamming indicator mainly quantifies radio interference in the GPS band that overpowers the standard information sent by the GNSS constellations.

The GPS jamming indicator varies from 0, which means no jamming, and goes up to 255, which means very intense jamming – see the green line in Fig. 1. Accordingly, with the PX4 documentation, to have no flight problems, GPS jamming indicator must be no more than 40. From my experience, in normal conditions the drone can still fly at values around 60 - 70 but no more than 90 without issues, but not over this value.

V. Results and Discussions

In order to have a baseline, referencing measurement for all subsequent analyses, the first study was carried out in the most unfavorable situation: no shielding at all and using both telemetry modules (the first line from Table II). Table II also shows the results obtained for EMI reduction by using the simplest possible method: we increased the distance between the interfering source and the receiver (i.e., the GSM module and magnetometer sensor).

Using a long mast decreases the mean GPS jamming indicator from an initial value of 79 to 64. So, using a longer support rod improved the overall electromagnetic characteristics of the system; in other words, there is less disturbance to the GPS and magnetometer systems. On the other side, from previous experiences [4], using a longer rod has also resulted in the amplification of GPS-magnetometer system vibrations and in a faulty yaw estimation. Hence, this method of EMI mitigation is not a very feasible one for our system.

The next logical step consisted of shielding, in turn, the Crossfire module and the 433 MHz telemetry module. Then, following an analysis of electromagnetic disturbances, we decided through which of the two components it is effective to send, in the future, the telemetry information.

The shielding used is very simple to implement. So, wrap the telemetry unit and all the cables between the FMU and the telemetry unit using food aluminum foil. In this mode, you will build a Faraday cage that will block almost all the telemetry unit's electromagnetic fields, see Fig. 2.

By using the telemetry data acquired only from the shielded Crossfire module (the 433 MHz module was entirely disconnected), the level of disturbances generated and recorded by the GSM module revealed a minimal value, of 27.23. Thus, the contribution of using the telemetry channel within the Crossfire module is almost zero. On the other hand, by using the telemetry channel on 433 MHz, very strong interference occurred that exceeded the maximum admissible threshold of 40, see Table III.

By using various SBCs together with a drone revealed the occurrence of numerous EMI problems [3], [4]. The logical solution to diminish the EMI generated by different onboard computers would be to shield the systems. But each shield has its effectiveness, known as shielding effectiveness. The shielding effectiveness is dependent on a large number of factors, such as the generated electromagnetic frequency spectrum, the electrical properties of the shielding material, the mechanical characteristic of the shield (e.g., shield thickness, the number and position of the discontinuities in the shielding material - such as access elements, connectors, or ventilation duct penetrations), etc. Even so, a subsystem, either shielded or unshielded, that generates less EMI will be more compatible with system's other components than any other similar subsystem producing more EMI. So, it is essential during designing and development stages to understand which SBC generates the specific electric and magnetic field interfering less with drone's GPS and magnetometer.

In order to obtain data from Table IV, on each development board was installed the latest version of the operating system (OS) recommended by the board producer at the time of writing of this text. For example, in the case of: (a) DragonBoard 410c (DB 410c), the Linaro Linux release 21.12 was used, (b) the i.MX 8M Mini (NavQ) ran Ubuntu 22.04 LTS, (c) on i.MX 8M Plus (NavQ Plus) is installed Ubuntu 20.04.5 LTS, (d) for the Raspberry Pi 3B+ (RPi 3 B+) and Raspberry Pi 4B (RPi 4) we tested four options – Raspberry Pi OS (based on Debian 11 - Bullseye) and Ubuntu (22.04.1 LTS), both on 32 and 64 bits –, (e) for Jetson Nano and Jetson Xavier NX the Linux4Tegra, a version of Ubuntu 18.04, was installed.

If present, the Bluetooth and Wi-Fi were deactivated on the systems presented in Table IV. More, 4 cores were used, the GUI has been disabled, and all systems have been programmed and configured from the command line interface (CLI). Finally, with all recordings, the same human detection program, based on a Single-Shot multibox Detection network, was executed on the same video file loaded from the SD card.

The first unexpected result was for DragonBoard 410c system. Even if this SBC was designed using EMI shields for SoC, memory, Wi-Fi, and Bluetooth, for RF noise-sensitive designs, the mean value of the GPS jamming parameter is quite large, taking a value of 50.

In the case of NavQ, the mean GPS jamming of 170 was huge. NavQ is built as a stack of three boards. From the construction, it comes with a camera, connected to a MIPI-CSI interface, and an HDMI convertor, connected to MIPI-DSI interfaces, with both components being connected to the second board. Due to the thinness of the connectors and the fear of destroying these boards, the first measurement was made with the video camera and HDMI interface connected. After removing the HDMI interface and the camera, the value decreased to around 80. Even if this investigation aims to analyze the EMI generated by SBCs, the presence of an HDMI video interface (with cables and connectors) gives a glimpse of other sources of interference. The HDMI is a well-known source of EMI [5].

The best board, from the electromagnetic compatibility point of view, is RPi 4, running the Raspberry Pi OS on 32 bits. However, when the Ubuntu 64-bit OS is used, the EMI shows a slight decrease compared to the situations of using Ubuntu 32-bit; this phenomenon is observed both at RPi 3B+ and RPi 4. The worst EMI case for the HoverGames drone is when the RPi 3 B+ board is used with the 32 bits Raspbian OS.

Also, a very interesting conclusion which can be drawn: running the same human detection algorithm on CUDA processors generates a minimum of EMI disturbances on the GPS unit, much less than running the same program on existing CPU units on the same development board, see Table IV. Moreover, even if the Jetson Xavier NX uses an SSD, which is considered a powerful source of EMI [1], the existence of the SSD does not seem to negatively or significantly impact the GPS.

A specific analysis was done for the NavQ Plus development board, and the results are presented in Table V. All these measurements were done in one session, composed of several different recordings, in the same mode as shown above in the "Database" section. Here, the RC command link was based on a TBS Crossfire 868 MHz long-range link, and the telemetry was sent back to the QGroundControl through a 433 MHz channel and system. I used a 433 MHz telemetry channel for these measurements to help people working with HoverGames drones, especially those participating in the “NXP HoverGames3: Land, Sky, Food Supply” contest.

The placement of the NavQ Plus development system relative to the other components of the drone is shown in Fig. 3.

As a starting point and to have a reference for the following determinations, a set of three EMI measurements were done without any development board or other electronic components placed on the drone other than the base ones required to fly the HoverGames drone (PDB, motors, ESCs, FMU, GPS and the telemetry unit) – the second line of the Table V. I know that comparing this result (70.79 mean GPS jamming) with the one presented in Table III fourth line is a very big difference (16.43 units) even if the almost all conditions were the same. The only justification I could find is that the recordings (those in Table III and those in Table V) were made in different sessions, on different days, but more importantly, in different places. So, in the new location where the measurements from Table V were made, the EMI environment was other (richer in EMI disturbances), and this is the cause why the results are so different.

Comparing the results generated by the NavQ (Table IV third line) and NavQ Plus(Table V) development systems, a real improvement in favor of the NavQ Plus system can be seen - keep in mind that the reference level is clearly against the NavQ Plus system. On the other hand, it should be noted that such a comparison is difficult to make since these systems generate disturbances that, in principle, cannot be said to be additive mainly because they may exist in different frequency bands.

Another interesting thing to note is that the NPU (Neural Processing Unit) generates much higher perturbations than the CPU when running the same algorithm to identify human subjects. It is also noted that the disturbances caused by the NavQ Plus development system alone (without considering other equipment that can be connected - such as a video camera) are significant and far exceed the limits allowed by the PX4 autopilot developers.

Another unexpected result is the substantial standard deviation of the obtained results - see Table V.

Fig. 4 shows the evolution over time of the GPS Jamming and GPS Noise parameters. This figure shows segments where (a) the GPS Jamming indicator has “reasonable” values in the range of 60, (b) time segments where the value of the parameter is around 110, and (c) short periods where this parameter takes values close to 250. I mention that during the entire recordings presented in Fig. 4, the recognition algorithm was running in the NPU of the NavQ Plus development system, and the GPS module was placed on the long rod.

All the results presented in Table IV and Table V were obtained in Hovering mode. But it is expected when the drone performs a real flight - with turns, level changes, or dives -, the minimum reference EMI value (27.23 for Table IV and 70.79 for Table V), used so far, to be completely different, mainly due to the power system's high voltage and current variations. In such a case, see Fig. 4(c), the computed mean GPS jamming is 47.55, and the corresponding standard deviation is 17.23; these data were obtained in the same measurement conditions as the ones used to get the result from Table III, second line. Also, when the drone is manually controlled (usually due to EMI jamming the GPS), the disturbances generated by the power elements are higher and the corresponding equivalent results in Table IV are expected to be slightly higher, too.

I also acknowledge that in a real EMI-rich environment (i.e., a city with 4G and 5G signals, with a lot of Wi-Fi stations, with hybrid and electric vehicles, etc.), the requirement for a threshold value for GPS jamming indicator less than 40 looks pretty restrictive; but, mainly because EMF levels cannot be accurately predicted, this value is only a conservative one.

VI. Conclusions

In conclusion, developing a drone able of real-time video monitoring a quarantine zone is a rather difficult task due to the various generated EMI components.

From the point of view of those developing intelligent algorithms using the existing neural processing kernel placed on the NavQ Plus system (with the help of NPU), methods to reduce the disturbances generated by the onboard system should undoubtedly be considered. Another mention to be made here is the need to use a long rod on which to place the GPS module to reduce the received disturbances, even if this option has disadvantages. One conclusion I came to it is that using this long rod brings more overall advantages than disadvantages.

Even if the Raspberry Pi 4 system achieves the optimal performance of EMI disturbances generated, it, unfortunately, lacks the computing power to implement complex, intelligent systems capable of working in real-time.

Although this study seems limited to the field of UAVs, its results can be used in many other related fields. Currently, most medical systems and measurement and control systems are intelligent ones, requiring the use of SBC systems. Therefore, realizing the magnitude of generated EMF, even only at a qualitative level, will allow the development of better and more robust professional systems.

If you wish to cite this work, please use the following reference:

Dan-Marius Dobrea, Monica-Claudia Dobrea, An L-Band EMI Analysis of Different Components Used with the NXP HoverGames UAV, E-Health and Bioengineering Conference (EHB), Nov. 17-18, 2022, România, Iași, eISSN:2575-5145, DOI: 10.1109/EHB55594.2022.9991300

References

[1] H.N. Lin, C.C. Lu, H.Y. Tsai, T.W. Kung, “The analysis of EMI noise coupling mechanism for GPS reception performance degradation from SSD/USB module”, International Symposium on Electromagnetic Compatibility, Tokyo, May 12-14, 2014

[2] L. Frezza, N. Picci, A. Gianfermo, E. Bedetti, D. Amadio, F. Curianò, F. Santoni, “LEDSAT 1U CubeSat GPS receiver electromagnetic interference (EMI) analysis”, 2021 IEEE 8th International Workshop on Metrology for AeroSpace (MetroAeroSpace), pp. 137-142, 2021

[3] M.C. Dobrea, D.M Dobrea, “A UAV development platform for intelligent applications, International Symposium on Signals, Circuits and Systems”, July 15-16, România, Iași, 2021

[4] D.M. Dobrea, M.C. Dobrea, “An autonomous UAV system for video monitoring of the quarantine zones”, Romanian Journal of Information Science and Technology, 23, no. S, 2020, pp. S53-S66

[5] C.Sreerama, “Effects of skew on EMI for HDMI connectors and cables”, International Symposium on Electromagnetic Compatibility, 2006