Improved Position, Navigation, and Timing

Accuracy for GPS-Aided INS in

GNSS Denied Environment

Understanding the Problem

For years Inertial Labs has produced high accuracyInertial Navigation Systems (INS) at the world’s bestprice-performance ratio. An INS estimates the position,attitude, and velocity using the gyroscopes andaccelerometers contained inside an inertialmeasurement unit (IMU). Position accuracy can begreatly improved when the INS is aided by GlobalNavigation Satellite System (GNSS). However, GNSS isnot always available. Customer requirements demandfor better performance of an INS during GNSS outage.Outages occur from natural environments like: tunnels,urban canyons, and roads under bridges; or from forcedconditions like GNSS jamming or spoofing. During GNSSoutage, position error growth is inevitable. Inertialnavigation using any IMU will accumulate error due tointegration and double integration of unknownaccelerometer and gyroscope biases. Using GNSS allowsthe INS to estimate and correct these biases. However,when satellite communication is lost, bias instability andnoise still cause drift. A typical INS using a MEMS-basedposition solution would drift by thousands of metersover 20 minutes of GNSS outage; a navigation-gradeIMU INS, like Northrop Grumman LN-100G, which useshigh accuracy ring laser gyros, claims to drift 120 metersafter a 20-minute GNSS outage. Instruments like this aresimply too large, too expensive, and too power hungryfor most applications, especially for the emerging ones.

Sample data was gathered to demonstrate theeffectiveness of Tunnel Guide for different applicationmethods in purely inertial mode, only using inertialsensors together with sophisticated modeling of landvehicle motion. For both of the following sample sets, adriving test was conducted using an Inertial Labs INS-Dwith Tunnel Guide without any aiding data from aDistance Measurement Instrument (DMI). The first testwas run on a highway driving at an average speed of 60mph (96 km/h). The plot below shows accumulated driftof approximately 160 meters, with outage time of 25minutes. This leads to an approximate drift ratecalculation of 0.4% of Distance Travelled (DT) for a 25-minute outage. This is already comparable to anavigation-grade INS (120-meter drift over 20-minuteoutage), yet the size, weight, power, and cost are anorder of magnitude lower.

The Tunnel Guide feature has proven to be an effectivesolution for military vehicles such as: Infantry CarrierVehicles (ICV), Mobile Gun Systems (MGS) Anti-TankGuided Missile Vehicles (ATGM), Anti-Drone GuidedMissile Vehicles (ADGMV), NBC Reconnaissance Vehicles(NBCRV), Assault Amphibious Vehicles (AAV), and LightArmored Multipurpose Vehicles (LAMV).

As a proven effective alternative, Inertial Labsdeveloped the Tunnel Guide feature for the GPS-AidedINS: an advanced algorithm that implements continuousdynamic modeling for land vehicle motion. As a result,the INS can mitigate error and increase accuracy of itsMEMS IMU during prolonged GNSS outage. Now theposition error of a land vehicle remains low even in aGNSS outage area such as: a tunnel; a city with signalblocking buildings; or any area where satellitecommunication is being blocked, spoofed, or jammed.

An Effective Solution

For applications involving more frequent turns, TunnelGuide will only improve in its accuracy.

The average drift rate of the second sample set wascalculated to be 0.2% DT over the course of the 30-minute outage with a peak drift of approximately 45meters. These results show that Tunnel Guidesignificantly improves performance accuracy duringGNSS outages, at a much lower cost.

A plot from another sample test was produced in asimilar manner, but in this situation the vehicle wasdriven in a city environment. This sample set shows theeffective nature of Tunnel Guide when vehicle dynamicsallows better observability of the system state. Thisdriving test was done at an average speed of 25 milesper hour (40 km/h) with frequent turns, and the GNSSoutage lasted for 30 minutes. The plot below showsperformance of the INS using Tunnel Guide with noaiding data from a DMI.

To test the filtered navigation solution while using aidingdata directly from a vehicle’s OBD-II port via CANbus, adrive test was performed in a similar manner to theprevious test. The vehicle was driven at an averagespeed of roughly 20 miles per hour, and a GNSS-outagewas simulated 15 minutes into the drive. After 30 moreminutes of driving under GNSS-outage conditions, thedevice was given GNSS corrections again, and thesolution converged. Above is an image of the test routethat was driven.

After the data was recorded and post-processed,Inertial Labs engineers derived a plot of position dataas a function of GNSS outage time. The derived plot isshown below.

Improved Accuracy with Aiding Data

Data was then processed in NovAtel Inertial Explorer.A post-processed solution utilized IMU data from theINS-B and the recorded GNSS data. This data wasindependent of the INS navigation solution and wasused as reference data for comparing the results of thesimulated GNSS-outage.

Additionally, a second plot was generated to show thetrendline for horizontal position error as a function ofdistance travelled. This plot is useful because it iscritical for land-based navigation systems that need tooperate in a GPS-denied environment. This plot isshown below.

In Summary

Regardless of your land vehicle application, keepingaccuracy during a GNSS outage can be the decidingfactor in meeting expectations of your projectrequirements. For Inertial Labs, the development of theTunnel Guide feature is a solution that providesincreased position accuracy without increasing the costof our devices. The GPS-aided INS-D is Inertial Labs’dual antenna INS unit, but with it comes accuracy andadvanced features that ensure that saving moneydoesn’t mean cutting corners on performance.

Specifications Accuracy (RMS)

Position with GNSS

1.5 m (SP)0.6 m (SBAS)0.4 m (DGPS)0.01 m (RTK)

Position without GNSS

0.2% DT (w/o odometer input)0.056% DT (w/ odometer input)

Pitch and Roll 0.05ᵒ

Heading/Yaw 0.08ᵒ

Velocity 0.03 m/sec

IMU1ᵒ/hr (Gyroscopes drift)

5 μg (Accelerometers drift)

Fluxgate Compass

Optional

Aiding DataOdometer, Encoder, Wheel

Sensor, DMI

Weight 320 grams

Size 120 x 50 x 53 mm

GPS-Aided INS-D

Trademark Legal Notice: All product names, logos, and brands are property of their respective owners.All company, product and service names used in this document are for identification purposes only. Useof names, logos, pictures and brands does not imply endorsement. Northrup Grumman and otherreferenced brands are trademarks of its affiliates or its respective owners.

From these plots, it can be concluded that the usage ofCANbus aiding data in the INS-D Kalman filter resultedin a horizontal position error of 0.056% DT, which is asignificant improvement on the 0.2% found in the testmentioned earlier. Now users can be confident thateven in an extended GPS-denied environment, systemscan navigate autonomously with accuracy thatoutperforms any other comparable sensor in theindustry.

What Do

You

Think?Here at Inertial

Labs, we care about our customers

satisfaction and want to continuously beable to provide solutions that arespecifically tailored to problems that areoccurring today, while vigorouslydeveloping products to tackle theproblems of tomorrow. Your opinion isalways important to us! Whether youare a student, an entrepreneur, or anindustry heavyweight. Share yourthoughts on our products,recommendations you have, or just sayhello at opinions@inertiallabs.com.

About Inertial Labs Inc.

Established in 2001, Inertial Labs is a leader in position and orientation technologies for commercial,

industrial, aerospace and defense applications. Inertial Labs has a worldwide distributor and

representative network covering 20+ countries across 6 continents and a standard product line spanning

from Inertial Measurement Units (IMU) to GPS-Aided Inertial Navigation Systems (INS). With application

breadth on Land, Air, and Sea; Inertial Labs covers the gambit of inertial technologies and solutions.

Inertial Labs, Inc.39959 Catoctin Ridge Street,Paeonian Springs, VA 20129 USAphone: +1 (703) 880 4222sales@inertiallabs.comwww.inertiallabs.com