Institute of Navigation

Institute of Navigation The world's premier professional society advancing positioning, navigation, and timing (PNT).

Heading to the Small Satellite Conference this month? The Institute of Navigation will be there August 23-26 in Salt Lak...
08/13/2026

Heading to the Small Satellite Conference this month?

The Institute of Navigation will be there August 23-26 in Salt Lake City at Booth 8.

Add us to your list. Stop by to talk PNT, the NAVIGATION journal, and what's ahead at ION GNSS+ 2026 in Orlando.

We'd love to meet you.

How long can a spacecraft keep watching a single pulsar before its own clock corrupts the measurement?As missions return...
08/12/2026

How long can a spacecraft keep watching a single pulsar before its own clock corrupts the measurement?

As missions return to the Moon and push deeper into space, navigation resources are stretched — the Deep Space Network is at times up to 40% oversubscribed. That has driven interest in autonomous alternatives like X-ray pulsar navigation (XNAV), which uses X-ray millisecond pulsars (MSPs) as signals of opportunity. But XNAV needs long integration times to form a usable measurement, and over those windows, error from the onboard clock accumulates unchecked.

Kyle J. Houser and Demoz Gebre-Egziabher (University of Minnesota) quantify how clock quality sets the limit on usable observation time. In a Monte Carlo simulation, a maximum-likelihood estimator (MLE) recovers phase and frequency from simulated photon arrival times corrupted by a two-state, frequency random walk clock error model. The estimator's error is compared against the Cramér–Rao lower bound (CRLB) across four oscillators — a TCXO, an OCXO, a Microchip SA65 chip-scale atomic clock (CSAC), and a cesium clock — using the pulsar PSR B1937+21.

In the authors' words: "How long can a pulsar be viewed without estimating or correcting clock errors before timing noise dominates the measurement?"

Each oscillator maps to a usable window: the TCXO is unsuitable, the OCXO holds for roughly 1,000–20,000 s, the CSAC stays converged to the CRLB until about 50,000 s, and the cesium clock is nearly indistinguishable from clean data. That distinction matters for small satellites, where a cesium clock is hard to fit within size, weight, power, and cost (SWAP-C) constraints.

Read the full open-access article in NAVIGATION:
https://doi.org/10.33012/navi.773

What happens to a UAV's navigation when GNSS is jammed and the vehicle is moving too fast for standard visual methods to...
08/05/2026

What happens to a UAV's navigation when GNSS is jammed and the vehicle is moving too fast for standard visual methods to keep up?

GNSS sensors are vulnerable to signal jamming and to the changing positions of satellites. Systems relying solely on an inertial measurement unit (IMU) accumulate positioning error that can reach the kilometer scale. Standard SLAM methods don't apply either: because the UAV moves at high speed along all three axes, the background continuously changes.

Ali Güven and İmam Şamil Yetik (TOBB Economics and Technology University, Ankara) present a camera-only navigation solution for continuously operating high-speed UAVs — the paper frames the problem as optimal route navigation for UAVs such as missiles or drones.

SuperPoint extracts features and LightGlue matches UAV camera images to reference satellite images. Two methods are introduced here: reference image optimization (RIO), which autonomously selects the reference images, and feature anomaly detection (FAD), which eliminates matched features inconsistent with camera movement. Between matching cycles, an XGBoost model tracks points using translation and rotation changes derived from homography estimation.

In the authors' words: "The results in Table 3 indicate that the proposed method does not exhibit the best tracking performance based on pixel error, but is very close to the best-performing method with a considerably lower computation time."

Evaluated entirely in simulation — Unreal Engine environments streamed via AirSim, plus Google Earth imagery — with runtimes measured on an Nvidia Jetson AGX Xavier. Reported target-based navigation error: 14.82 ± 2.42 m.

Read the full open-access article in NAVIGATION:
https://lnkd.in/eUYabKeW

When you're validating a high-precision GNSS solution, how much should you trust the resolved integer ambiguities — espe...
07/29/2026

When you're validating a high-precision GNSS solution, how much should you trust the resolved integer ambiguities — especially when you're not fully confident they're correct?

Model validation is what catches the unmodeled effects — a blunder in a pseudorange, a carrier-phase cycle slip, or an atmosphere delay that isn't fully corrected — that bias the estimated parameters, such as user position coordinates, when they remain unnoticed. Two common approaches each have a catch. The ambiguity-float (AF) approach ignores the integer nature of the ambiguities, so validation can't benefit from those constraints. The alternative assumes the ambiguities are known, which only holds when the resolution success rate is very close to one.

Chengyu Yin, Peter Teunissen, and Christian Tiberius (Delft University of Technology, with Teunissen also at Melbourne and Curtin) take up a middle path. Their paper shows how to apply two AR-based parameter significance tests — the ARs test, which uses a high-probability-density acceptance region, and the ARn test, which uses a continuous ellipsoidal one — both of which use the distribution of the resolved ambiguities. They evaluate detection power against the AF and ambiguity-known (AK) tests across blunders, cycle slips, and ionosphere and troposphere delays.

In the authors' words: "the AR significance tests can perform better than the AF test, even if the success rate is not close to one."

The results come from Monte Carlo experiments on a single-constellation GPS double-differenced model. The advantage isn't universal — the AR test can sometimes trail the AF test, so the authors give an easy-to-compute criterion for the blunders and cycle slips tests to predict which case you're in. For atmosphere delays, where the power functions turn spiky, they show that a partial ARs (PARs) test — resolving a subset of the ambiguities — keeps the power function smooth, and that combining the ARs and PARs tests always outperforms either alone.

Read the full open-access article in NAVIGATION: https://lnkd.in/eFwCG8vT

How does a self-driving car stay in its lane when the buildings around it are scrambling its satellite signals?In dense ...
07/22/2026

How does a self-driving car stay in its lane when the buildings around it are scrambling its satellite signals?

In dense cities, GNSS signals reflect off building faces before reaching the receiver, corrupting the measurements a vehicle relies on for absolute positioning. Autonomous driving demands tight localization — the paper cites a horizontal 95% accuracy requirement of 0.21 m for autonomous vehicle operations (Reid et al.). And when more than half the satellites in view are non-line-of-sight (NLOS), conventional robust methods reach their breakdown point and the position solution degrades.

Fabian Ruwisch and Steffen Schön of the Institut für Erdmessung (Institute of Geodesy) at Leibniz Universität Hannover address this by pairing stored map information with robust estimation for GNSS RTK positioning.

Their approach introduces the HG-estimator, a new robust estimator combining the Huber and Geman–McClure loss functions. It is embedded in GNSS feature map-aided weighting (GNSS FMA-W), where a GNSS feature map supplies a prior on each satellite's expected pseudorange residual along the route. That prior lets the weight model down-weight corrupted signals rather than excluding them outright.

The method was validated on two real automotive test drives in Hannover — a medium and a deep urban trench with line-of-sight satellite availability below 50%. Combining the GNSS feature map with the HG-estimator improved horizontal accuracy by 54% and 60% over C/N0 weighting respectively, and was the only tested method to meet lane-determination requirements in the deep trench.

Read the full open-access article in NAVIGATION: https://doi.org/10.33012/navi.779

Happy Independence Day from everyone at the Institute of Navigation!As millions travel by road, air, and sea this Fourth...
07/03/2026

Happy Independence Day from everyone at the Institute of Navigation!

As millions travel by road, air, and sea this Fourth of July to be with family and friends, it's worth a quiet nod to the positioning, navigation, and timing technology quietly guiding the way.

Wishing our members celebrating the holiday a safe, happy holiday.

What happens to a vehicle's navigation system when the GNSS signal disappears — underground, underwater, in space, or ac...
06/24/2026

What happens to a vehicle's navigation system when the GNSS signal disappears — underground, underwater, in space, or across military and aerospace operations?

Inertial navigation systems estimate position, velocity, and attitude from accelerometers and gyroscopes alone, with no external infrastructure. But they accumulate drift over time, and real-world conditions introduce non-Gaussian noise, sensor nonlinearities, and unmodeled dynamics that standard filters handle poorly.

Farnaz Imani and Jafar keighobadi (University of Tabriz) take this on. Their study presents INS-ATOMIC, an inertial navigation system that runs entirely on atomic accelerometers and gyroscopes with no GNSS aiding in the filtering loop. They compare two robust filtering strategies: the chi-square increment robust Kalman filter (CI-RKF), which rejects statistically inconsistent measurements, and the multi-kernel maximum correntropy extended Kalman filter (MKMC-EKF), which uses multiple kernel functions of differing bandwidths to downweight outliers and heavy-tailed noise.

In the authors' words: "The primary innovation of this study lies in the practical application of the MKMC-EKF in conjunction with atomic sensors under real-world conditions."

The filters were evaluated on a single test vehicle driven along a campus route at the University of Tabriz, with VITANS reference measurements (embedding GPS) used only to generate ground-truth trajectories for error analysis. Relative to the CI-RKF, the paper reports root-mean-square-error reductions from 37% to over 99% in velocity, and up to 99% in acceleration and angular rates.

Read the full open-access article in NAVIGATION: https://lnkd.in/ePTvYQH5

What happens to navigation signals when they are transmitted from fast-moving low-Earth-orbit satellites through a distu...
06/17/2026

What happens to navigation signals when they are transmitted from fast-moving low-Earth-orbit satellites through a disturbed ionosphere?

As interest grows in using LEO satellites for future positioning, navigation, and timing, one challenge is understanding how ionospheric scintillation affects signals transmitted from these highly dynamic platforms — especially across different frequency bands.

A new paper in NAVIGATION by Dr. Andrew K. Sun, Prof. Jade Morton, Dr. Charles Rino, and Prof. Jiyun Lee examines ionospheric scintillation effects on LEO-transmitted signals across VHF, UHF, L-band, and S-band frequencies.

Using a physics-based multi-frequency scintillation simulator, the study compares three signal-dynamics scenarios: conventional MEO transmission, a lower-dynamics LEO case, and a higher-dynamics LEO case. The simulations show that higher LEO signal dynamics compress scintillation effects in time, producing more frequent signal fades and larger phase-rate variations.

Frequency also matters. Lower-frequency signals experience stronger scintillation effects, including more frequent and deeper fades and larger phase fluctuations. In the most dynamic LEO case, VHF signals exhibit extremely rapid fading, while S-band signals show substantially fewer deep fades.

A key quantitative finding is that the temporal characteristics of scintillation — including fade time separation and fade duration — scale consistently with the time-to-space scintillation parameter across frequency bands. This provides a useful way to relate scintillation behavior under different LEO signal-dynamics conditions.

The paper also finds that simultaneous deep fades across L1, L2, and L5 remain relatively rare, even under high LEO dynamics. This suggests that inter-frequency aiding and frequency diversity may help improve receiver tracking robustness for future LEO-based navigation systems.

Read the full paper, "Ionospheric Scintillation Effects on LEO-Transmitted Signals Across Multiple Frequency Bands" on our open-access website: https://navi.ion.org/content/73/1/navi.772/tab-article

JNC 2026 is in the books.Four days, all of it focused on advancing positioning, navigation, and timing for those who ser...
06/05/2026

JNC 2026 is in the books.

Four days, all of it focused on advancing positioning, navigation, and timing for those who serve.

A big thank you to the ION Military Division leadership, speakers and session chairs. They guide this conference and give their time to keep the PNT community moving forward.

Grateful for everyone who made the trip. See you next year.

06/02/2026

The exhibit hall is open and the week is off to a strong start.

Military leaders, government agencies, industry professionals, and researchers coming together to discuss the latest developments in positioning, navigation, and timing (PNT).

Welcome to JNC 2026.

Address

8551 Rixlew Lane, Ste 360
Manassas, VA
20109

Opening Hours

Monday 9am - 5pm
Tuesday 9am - 5pm
Wednesday 9am - 5pm
Thursday 9am - 5pm
Friday 9am - 5pm

Website

Alerts

Be the first to know and let us send you an email when Institute of Navigation posts news and promotions. Your email address will not be used for any other purpose, and you can unsubscribe at any time.

Shortcuts

Share

Category