Simultaneous Tracking and Navigation (STAN): UAV Navigation with LEO Satellite Signals

TThis video presents an experimental demonstration of an unmanned aerial vehicle's (UAV) inertial navigation system (INS) being aided by low Earth orbit (LEO) satellite signals. While GNSS signals are available, the UAV uses the INS aided by GNSS and LEO satellite signals to navigate while simultaneously tracking the positions of multiple LEO satellites. When GNSS signals become unavailable, the UAV performs simultaneous tracking and navigation (STAN), i.e., navigates exclusively with the INS aided by LEO satellite signals while continuing to simultaneously track the LEO satellites. An improved LEO satellite dynamics model (two-body plus J2) is used in the navigation filter. Received signals from 2 Orbcomm LEO satellites were processed with the Multichannel Adaptive TRansciever Information eXtractor (MATRIX) software-defined receiver (SDR) developed by the ASPIN Laboratory to produce Doppler measurements. The ground truth trajectory of the UAV was obtained by parsing the navigation solution from the drone's onboard navigation system, which consists of a multi-constellation GNSS receiver, an INS, and an altimeter. Two approaches are compared: (1) INS only and (2) STAN using a two-body plus J2 model for LEO satellite position and velocity propagation. The UAV navigated over a 1.4 km trajectory in 2 minutes and 35 seconds, during which GNSS signals were unavailable for the last 45 seconds. The following navigation results were achieved for the two approaches: (1) final position error of 123.5 m and position root mean squared error (RMSE) of 53.7 m and (2) final position error of 5.7 m and position RMSE of 5.4 m.

 

Simultaneous Tracking and Navigation (STAN): UAV Navigation with LEO Satellite Signals
 
 

Sub-Meter Accurate UAV Navigation and Cycle Slip Detection with LTE Carrier Phase Measurements

This video shows an unmanned aerial vehicle (UAV) navigating with cellular long-term evolution (LTE) signals with a sub-meter level accuracy. The UAV navigates with the Multichannel Adaptive TRansceiver Information eXtractor (MATRIX) software-defined receiver (SDR), developed by the ASPIN Laboratory, which produces carrier phase, code phase, and Doppler frequency measurements from received LTE signals. Single difference measurements are used to remove the effect of the receiver’s clock bias. LTE eNodeBs’ clock biases are initialized using the known initial position of the UAV. The measurements are fused via an extended Kalman filter (EKF) to estimate the UAV’s position and integer ambiguities of the carrier phase single difference measurements. The MATRIX SDR uses carrier phase measurements from LTE eNodeB’s multiple antenna ports to detect cycle slips. The UAV traverses a trajectory of 605 m in 175 seconds, achieving a two-dimensional (2-D) position root-mean-squared error (RMSE) of 81 cm between (1) the LTE navigation solution and (2) the ground truth navigation solution obtained from a global navigation satellite system (GNSS) receiver with real-time kinematic (RTK) coupled with an inertial measurement unit (IMU).

 

Sub-Meter Accurate UAV Navigation and Cycle Slip Detection with LTE Carrier Phase Measurements
 
 

Localization using Differential Carrier Phase Measurements from Orbcomm LEO Satellite Signals

This video presents a ground receiver localizing itself with Orbcomm low Earth orbit (LEO) satellite signals. The receiver makes carrier phase measurements to Orbcomm LEO satellites, and fuses its measurements with carrier phase measurements from a reference receiver through a carrier phase differential (CD)-LEO framework. The reference receiver is mounted on an unmanned aerial vehicle (UAV), which had knowledge of its position from GNSS signals. The ground receiver, located near the reference UAV, does not know its position and is using the CD-LEO framework for localizing itself. The ground receiver was able to localize itself to within less than 12 m after 114 seconds, by listening to only 2 Orbcomm satellites.

 

Localization using Differential Carrier Phase Measurements from Orbcomm LEO Satellite Signals
 
 

UAV Integrity Monitoring Measure improvement using Terrestrial Signals of Opportunity

This video presents an innovative approach to improve the integrity measures of an unmanned aerial vehicle (UAV) by exploiting ambient terrestrial signals of opportunity (e.g., cellular, digital television, and AM/FM radio signals). Integrity refers to the system’s ability to detect anomalies and warn the user about when the system should not be used. Experimental results to evaluate the performance of the developed approach shows a vertical protection improvement of nearly 60% and a horizontal protection improvement of about 76%, compared to GPS only.

 

UAV Integrity Monitoring Measure improvement using Terrestrial Signals of Opportunity
 
 

A Pedestrian Indoor Localization System Using LTE Signals and Synthetic Aperture Navigation

This video demonstrates a pedestrian indoor navigation system that uses long-term evolution (LTE) signals in a synthetic aperture navigation (SAN) framework. The proposed framework exploits the motion of the receiver to synthesize an antenna array from time-separated elements. The synthesized data received by the synthetic antenna array is processed to suppress multipath via determining the direction-of-arrival (DOA) of incoming signals. The proposed LTE-SAN framework is compared with a state-of-the-art standalone LTE navigation receiver. The video shows a pedestrian navigating in a multipath-rich indoor environment while listening to 6 LTE eNodeBs, traversing a trajectory of 126.8 m, and achieving a two-dimensional position root mean-squared error (RMSE) of 3.93 m.

 

A Pedestrian Indoor Localization System Using LTE Signals and Synthetic Aperture Navigation
 
 

LEO Satellite Signal-Aided Inertial Navigation with Periodically Transmitted Satellite Positions

This video presents an experimental demonstration of a ground vehicle's inertial navigation system (INS) being aided by low Earth orbit (LEO) satellite signals. While GNSS signals are available, the ground vehicle uses the INS aided by GNSS and LEO satellite signals to navigate while simultaneously tracking the positions of multiple LEO satellites. When GNSS signals become unavailable, the ground vehicle performs simultaneous tracking and navigation (STAN), i.e., navigates exclusively with the INS aided by LEO satellite signals while continuing to simultaneously track the LEO satellites. Received signals from 2 Orbcomm LEO satellites were processed with the Multichannel Adaptive TRansciever Information eXtractor (MATRIX) software-defined receiver (SDR) developed by the ASPIN Laboratory to produce Doppler measurements. The ground truth trajectory of the vehicle was obtained with a multi-frequency, dual antenna GNSS receiver with real-time kinematic (RTK) and a tactical grade inertial measurement unit (IMU). Three approaches are compared (1) INS only; (2) STAN without decoding the LEO satellites' positions (which happen to be periodically transmitted by the LEO Orbcomm satellites); and (3) STAN with decoding these positions and including them in the navigation filter. The results show the ground vehicle navigating over a 7.5 km trajectory in 4 minutes and 18 seconds, during which GNSS signals were only available for the first 30 seconds. The following results were achieved with the three approaches: (1) final position error of 16,589 m and position root mean squared error (RMSE) of 6,864.6 m; (2) final position error of of 476.3 m and position RMSE of 195.6 m; (3) final position error of 233.3 m and position RMSE of 188.6 m.

 

LEO Satellite Signal-Aided Inertial Navigation with Periodically Transmitted Satellite Positions
 
 

Event-Based Distributed Cellular-Aided Inertial Navigation

This video presents an experimental demonstration of an event-based distributed signal of opportunity (SOP)-aided inertial navigation system (INS) framework. Multiple unmanned aerial vehicles (UAVs) transmit a packet to each other containing INS information and pseudorange observables extracted from cellular SOP transmitters, which are fused to aid the UAVs' on-board INSs. Instead of transmitting the packet at a fixed communication rate, an event-based strategy is used to communicate this packet, where the communication event is triggered whenever the probability of maximum position error (p) could violate a desired threshold (e). While GPS is available, the UAVs use their INSs aided by GPS and cellular SOPs to navigate, while simultaneously mapping the cellular transmitters. When GPS signals become unavailable, the UAVs navigate exclusively with their INSs aided by cellular SOP pseudoranges, while simultaneously mapping the cellular transmitters (i.e., performing collaborative radio SLAM). Results demonstrate that the event-based communication strategy (with p = 0.999 and e = 10 m) reduces the amount of communicated data by 77.3% compared to a fixed-rate communication strategy.

 

Positioning with Cellular LTE Signals Exploiting Antenna Motion
 
 

Positioning with Cellular LTE Signals Exploiting Antenna Motion

This video presents a ground vehicle navigating with cellular long-term evolution (LTE) signals and inertial measurement unit (IMU) measurements. The LTE measurements were obtained using the Multichannel Adaptive TRansceiver Information eXtractor (MATRIX) software-defined receiver (SDR), developed by the ASPIN Laboratory. The navigation solution obtained by our proposed approach is based on a multi-state constraint Kalman filter (MSCKF). The results show that the proposed approach are more accurate than a navigation solution which fuses LTE signals with IMU measurements through a standard extended Kalman filter (EKF). This is due to the fact that multipath introduces time-correlated errors in LTE pseudoranges, which are not accounted for in a standard EKF, while an MSCKF accounts for them. The ground truth trajectory of the vehicle was obtained with a multi-frequency, dual antenna GNSS receiver with real-time kinematic (RTK) and a tactical grade IMU. The proposed approach reduces the 2D and 3D position root mean squared-error (RMSE) by 29% and 64.7%, respectively, over the EKF; and the 2D and 3D maximum error by 19.6% and 86.7% respectively, over the EKF.

 

Positioning with Cellular LTE Signals Exploiting Antenna Motion
 
 

Centimeter-Accurate UAV Navigation with Carrier Phase Differential-Cellular Measurements

This video shows an unmanned aerial vehicle (UAV) navigating exclusively using cellular signals. The cellular carrier phase measurements were obtained using the Multichannel Adaptive TRansceiver Information eXtractor (MATRIX) software-defined receiver (SDR), developed by the ASPIN Laboratory. The navigation framework includes a base in the UAV's vicinity, enabling the UAV to navigate with carrier phase differential cellular (CD-cellular) signals from nearby base transceiver stations (BTSs). The navigation framework requires no prior knowledge of the UAV's position and achieves centimeter-level positioning accuracy. A batch weighted nonlinear least-squares estimator is used to solve for the integer ambiguities, and an extended Kalman filter is used to initialize the batch estimator. The video shows the UAV navigating over a 1.72 km trajectory in 3 minutes with 62.11 cm root mean-square error (RMSE).

 

Centimeter-Accurate UAV Navigation with Carrier Phase Differential-Cellular Measurements
 
 

Indoor Navigation System with an LTE-Aided Inertial Measurement Unit

This video demonstrates an indoor pedestrian navigation system that uses long-term evolution (LTE) signals and an inertial measurement unit (IMU). The navigation system fuses LTE signals and the IMU in a tightly-coupled fashion via an extended Kalman filter. The LTE carrier phase measurements were obtained using the Multichannel Adaptive TRansceiver Information eXtractor (MATRIX) software-defined receiver (SDR), developed by the ASPIN Laboratory, and the IMU measurements were obtained using a tactical-grade IMU. The indoor navigation framework employs an outdoor receiver, referred to as the base, which estimates the unknown clock biases of LTE eNodeBs and shares these estimates with the indoor navigating receiver. The video shows a pedestrian navigating indoors over a trajectory of 109 m with a two-dimensional position root mean-squared error (RMSE) of 2.92 m.

 

Indoor Navigation System with an LTE-Aided Inertial Measurement Unit
 
 

Integrity Monitoring of LTE Signals of Opportunity-Based Navigation for Autonomous Ground Vehicles

This video demonstrates a receiver autonomous integrity monitoring (RAIM) framework for autonomous ground vehicle navigation by fusing an inertial measurement unit (IMU) with cellular long-term evolution (LTE) signals, in the absence of GNSS signals. The received cellular LTE signals were processed with the Multichannel Adaptive TRansciever Information eXtractor (MATRIX) software-defined receiver (SDR) developed by the ASPIN Laboratory. The RAIM framework employs a fault detection test to identify and exclude faulty LTE measurements. This video shows a ground vehicle navigating via this framework in an urban environment (downtown Riverside, California) over a trajectory of 1.15 km. It is demonstrated that the RAIM framework detects and excludes faulty LTE pseudorange measurements, reducing the position root mean-squared error (RMSE) by 66.2% from the case where such faulty pseudorange measurements are not excluded.

 

Integrity Monitoring of LTE Signals of Opportunity-Based Navigation for Autonomous Ground Vehicles
 
 

LEO Satellite Signal-Aided Inertial Navigation

This video presents the firstexperimental demonstration of an unmanned aerial vehicle's (UAV's) inertial navigation system (INS) being aided by low Earth orbit (LEO) satellite signals. While GNSS signals are available, the UAV uses the INS aided by GNSS and LEO satellite signals to navigate while simultaneously tracking the positions of multiple LEO satellites. When GNSS signals become unavailable, the UAV performs simultaneous tracking and navigation (STAN), i.e., navigates exclusively with the INS aided by LEO satellite signals while continuing to simultaneously track the LEO satellites. Received signals from 2 Orbcomm LEO satellites were processed with the Multichannel Adaptive TRansciever Information eXtractor (MATRIX) software-defined receiver (SDR) developed by the ASPIN Laboratory to produce Doppler measurements. The ground truth trajectory of the UAV was obtained with a multi-frequency, dual antenna GNSS receiver with real-time kinematic (RTK) and a tactical grade inertial measurement unit (IMU). Two approaches are compared (1) INS only and (2) STAN. The results show the UAV navigating for 120 seconds, during which GNSS signals were only available for the first 90 seconds. The following results were achieved with the two approaches: (1) final position error of 31.7 m and position root mean squared error (RMSE) of 14.4 m and (2) final position error of of 8.8 m and position RMSE of 6.8 m.

 

LEO Satellite Signal-Aided Inertial Navigation
 
 

Pseudorange and Multipath Analysis of Positioning with LTE Secondary Synchronization Signals

This video presents a ground vehicle navigating with cellular long-term evolution (LTE) secondary synchronization signal (SSS) in an urban multipath environment: downtown Riverside, California. These results were obtained using the Multichannel Adaptive TRansceiver Information eXtractor (MATRIX) LTE SDR developed in the ASPIN Laboratory. The pseudorange error statistics, which were derived analytically, were used in a weighted nonlinear least squares (WNLS) framework to estimate the position of the ground vehicle. The results demonstrate a reduction of 51% in the root-mean squared error (RMSE) and 21% in the maximum error when compared to a nonlinear least squares (NLS) framework that does not weight the pseudorange measurements.

 

Pseudorange and Multipath Analysis of Positioning with LTE SSS
 
 

 

Game of Drones 2018 - Episode 3

This video shows Episode 3 of ASPIN Laboratory's annual Game of Drones: an annual summer event in which middle and high school students are invited to participate in a game of drones. These games span half a day of fun and engaging games involving drones to teach and inspire students to pursue science, technology, engineering, and mathematics (STEM). A total of 81 students participated in Episode 3 of Game of Drone. ASPIN Laboratory has been organizing and hosting these games since 2016.

 

Game of Drones 2018 - Episode 3
 
 

 

SLAM with Cellular-Aided Lidar Measurements for Lane-Level Accurate Ground Vehicle Navigation

This video presents lane-level accurate ground vehicle navigation by fusing Lidar with cellular long-term evolution (LTE) signals, without GPS signals. The received cellular LTE signals were processed with the Multichannel Adaptive TRansciever Information eXtractor (MATRIX) software-defined receiver (SDR) developed by the ASPIN Laboratory. The vehicle simultaneously estimates its own pose (three-dimensional position and orientation), the LTE towers’ three-dimensional position, and the difference between the vehicle's receiver clock bias and drift and the bias and drift of each LTE tower. This framework uses a minimal number of the lidar point cloud to produce a navigation solution, which makes it suitable for real-time implementations. In this video, only 4.6% of the points returned by the lidar were used.

 

Collaborative Cellular-Aided Inertial Navigation
 
 

 

Submeter-Accurate UAV Navigation with Cellular CDMA Carrier Phase

This video presents an experimental demonstration of a UAV navigating exclusively with cellular CDMA signals using precise carrier phase measurements. The UAV was listening to 9 cellular CDMA base transceiver stations (BTSs). While the UAV has access to GPS, it characterizes the relative frequency stability of the BTSs then intelligently clusters them. This clustering minimizes the error in the position estimate obtained exclusively via cellular CDMA signals. The received cellular signals were processed with the Multichannel Adaptive TRansciever Information eXtractor (MATRIX) software-defined receiver (SDR) developed by the ASPIN Laboratory. The cellular CDMA navigation solution is compared to the solution obtained from the UAV's on-board navigation system (the A3 flight controller), which consists of a GPS receiver, an IMU, a barometer, and a magnetometer. The cellular CDMA navigation solution over more than 5 minutes of flight time was within 88.58 cm from the GPS navigation solution.

 
 
 
Collaborative Cellular-Aided Inertial Navigation 

 

Distributed Cellular-Aided Inertial Navigation with Intermittent Communication

This video presents an experimental demonstration of a distributed signal of opportunity (SOP)-aided inertial navigation system (INS) framework. This framework enables multiple unmanned aerial vehicles (UAVs) to aid their on-board INSs by communicating INS information and mutual pseudorange observables extracted from cellular transmitters.  The pseudoranges were obtained using the Multichannel Adaptive TRansceiver Information eXtractor (MATRIX) LTE SDR developed in the ASPIN Laboratory. This framework is studied in a lossy wireless communication channel with a probability of packet drop p. While GPS is available, the UAVs use their INSs aided by GPS and cellular signals to navigate, while simultaneously mapping the cellular transmitters. When GPS signals become unavailable, the UAVs navigate exclusively with their INSs aided by cellular pseudoranges, while simultaneously mapping the cellular transmitters (i.e., performing collaborative radio SLAM). Results demonstrate that even with a high probability of packet drop (p = 0.7), the exploitation of free ambient cellular SOPs in the environment significantly reduces INS errors in the absence of GPS. The results are compared with a perfect communication assumption (p = 0).

 
 
 
 
Collaborative Cellular-Aided Inertial Navigation

 

Ground Vehicle Navigation with LTE Signals in a Multipath Environment

This video presents ground vehicle navigation using long-term evolution (LTE) signals in an urban multipath environment: downtown Riverside, California. These results were obtained using the Multichannel Adaptive TRansceiver Information eXtractor (MATRIX) LTE SDR developed in the ASPIN Laboratory. This computationally efficient receiver exploits the cell-specific reference signal (CRS) to estimate the time-of-arrival. The high transmission bandwidth of the CRS (up to 20 MHz) makes it robust in a multipath environment. Although the received LTE signals experience more multipath than GPS signals due to the low elevation angles at which signals are received, the results show meter-level accuracy in navigation with the MATRIX LTE SDR compared to the GPS navigation solution.

 
 
Collaborative Cellular-Aided Inertial Navigation

 

Collaborative Cellular-Aided Inertial Navigation

This video presents the first experimental demonstration of multiple unmanned aerial vehicles (UAVs) aiding their on-board inertial navigation systems (INSs) by sharing mutual pseudorange observables extracted from cellular transmitters and communicating their own inertial measurement unit (IMU) data. While GPS is available, the UAVs use their INSs aided by GPS and cellular signals to navigate while simultaneously mapping the cellular transmitters. When GPS signals become unavailable, the UAVs navigate exclusively with their INSs aided by cellular pseudoranges while simultaneously mapping the cellular transmitters (i.e., performing centralized collaborative radio SLAM). Results demonstrate that the exploitation of free ambient cellular signals of opportunity (SOPs) in the environment significantly reduces INS errors in the absence of GPS.

 
Collaborative Cellular-Aided Inertial Navigation

 

Ground Vehicle Navigation with LTE Signals: SSS vs. CRS

This video presents ground vehicle navigation using two different reference signals in a semi-urban environment in long-term evolution (LTE) systems: the secondary synchronization signal (SSS) and the cell-specific reference signal (CRS). The transmission bandwidth of the SSS is less than 1 MHz, leading to low time-of-arrival (TOA) estimation accuracy in a multipath environment. The CRS is more robust in multipath environments due to its higher transmission bandwidth, which can be as high as 20 MHz. The navigation solutions estimated from the SSS only and from the SSS aided by the CRS are shown. For the SSS-only solution, a computationally-efficient receiver was designed. For the CRS-aided SSS solution, the channel impulse response was estimated using the CRS and used as a feedback into the SSS receiver tracking loops. The results show 5 times improvement in the root mean squared error (RMSE) by using the CRS-aided SSS receiver over the SSS receiver.

 
 
Ground Vehicle Navigation with LTE Signals: SSS vs. CRS 

 

Cellular-Aided Inertial Navigation

This video presents the first experimental demonstration of an unmanned aerial vehicle's (UAV's) inertial navigation system (INS) being aided by a cellular signal of opportunity (SOP). While GPS is available, the UAV uses the INS aided by GPS and cellular signals to navigate while simultaneously mapping the cellular SOP. When GPS signals become unavailable, the UAV navigates exclusively with the INS aided by the cellular SOP while simultaneously mapping the cellular SOP (i.e., performing radio SLAM). Results demonstrate that the exploitation of free ambient cellular signals in the environment significantly reduces INS errors in the absence of GPS and bounds the INS drift.

Cellular-Aided Inertial Navigation

 

UAV Navigating with Cellular LTE Signals

This video presents the first experimental demonstration of a UAV navigating exclusively with cellular LTE signals. In this video, a DJI Matrice 600 is equipped with a cellular consumer-grade 800/1900 MHz omnidirectional antenna for receiving LTE signals. The LTE signals were down-mixed and sampled via an Ettus E312 USRP. Then, the signals were processed using the LTE software-defined receiver (SDR) developed by ASPIN Laboratory. The LTE navigation solution is compared to the UAV's true trajectory, which is obtained from the UAV's sensor suite (e.g., GPS, IMU, barometer, etc.).

UAV Navigating with Cellular LTE Signals

 

UAV Navigating with Cellular CDMA Signals

This video presents the first experimental demonstration of a UAV navigating exclusively with cellular CDMA signals. In this video, a DJI Matrice 600 is equipped with a cellular consumer-grade 800/1900 MHz omnidirectional antenna for receiving CDMA signals. The CDMA signals were down-mixed and sampled via an Ettus E312 USRP. Then, the signals were processed using the LabVIEW-based cellular CDMA software-defined receiver (SDR) developed by ASPIN Laboratory. The UAV is receiving information about ambient cellular signals from a reference station deployed in the University of California, Riverside campus. The cellular CDMA navigation solution is compared to the solution obtained from the UAV's on-board navigation system (the A3 flight controller), which consists of a GPS receiver, an IMU, a barometer, and a magnetometer.

  UAV Navigating with Cellular CDMA Signals

 

GPS Vertical Dilution of Precision Reduction using Signals of Opportunity

This video presents simulated and experimental demonstrations of exploiting signals of opportunity (SOPs), e.g., cellular signals, to reduce the relatively large vertical errors that are intrinsic to a GPS-only navigation solution. Since cellular signals are transmitted from towers located in favorable geometric configurations, fusing them with GPS signals significantly improves the accuracy when compared to a GPS-only navigation solution.