G1-m1TM Static White Paper.docx rev 1.0, October 2013

G1-m1TM GNSS Mobile System

Field Testing - Static Case

WWW.GEOMATICS.US

October 2013

© 2013 Geomatics USA, LLC All Rights Reserved

Purpose: This document is for the G1 GNSS system users.

It shows the achieved accuracy of the G1-m1TM system

data in static mode. For instructions on how to acquire

the GNSS data, refer to the G1 receiver quick guide; for

instructions on how to process the acquired data, refer to

the HorusTM software guide.

Requirements: This document presumes that you have

at least one G1 GNSS receiver (the rover). Optionally,

users can collect correction data on a known base station

(the base). Or, you can request to download the

correction data from the closest Continuously Operating

Reference Station (CORS), as described in the HorusTM

software guide.

Note: By GNSS data, we mean raw ranges between the

G1 receiver and the respective GPS/GLONASS satellites in

the form of C/A code and carrier phase observations. The

raw ranges are readily available on a uSD card in the form

of a binary data file after a successful data acquisition

mission using the G1 receive.

Test Objective

The G1-m1TM receiver has an L1-only GNSS engine. It is

capable of acquiring data in static and kinematic

positioning modes. This test is to show the achievable

positioning accuracy of the G1-m1TM system in static

mode, i.e. when the system is stationary over a point for

a while. The G1-m1TM system used is configured for 1-s

interval and GPS-only mode. The antenna used is a G1-

A311TM patch with a 3.75” ground plane.

Figure 1: G1-m1TM System on a mini Tripod over a GCP Check Point

G1-m1TM Field Testing – Static Case

Test Site

The test was conducted in June 2013 at the parking lot of

the Santa Fe Community College in Gainesville, FL. The

site was chosen because of its relative openness and its

suitable satellite visibility, see Figure 2.

Reference Network

A network of three control stations were surveyed with

Trimble 5700 geodetic-grade receivers to establish a

reference for the survey, as shown in Figure 3.

The reference network was adjusted using Trimble

Business Center. The summary of the network adjustment

report is shown in Table 1. The reference network

provides a benchmark solution accurate to a few

millimeters both horizontally and vertically. For the

purpose of this evaluation, we shall consider the reference

network stations to be fixed (errorless.)

GNSS Loop Closure Results Summary

Legs in Loop = 3 Length

(Meter)

ΔHoriz

(Meter)

ΔVert

(Meter) PPM

Pass/Fail Criteria 1

Best 0.001 -0.003 5.743

Worst 0.001 -0.003 5.743

Average Loop 590.814 0.001 0.003 5.743

Standard Error 0.000 0.007 0.000 0.000

Adjustment Statistics

Number of Iterations for Successful Adjustment: 2

Network Reference Factor: 1.00

Chi Square Test (95%): Passed

Precision Confidence Level: 95%

Degrees of Freedom: 3

Post Processed Vector Statistics

Reference Factor: 1.00

Redundancy Number: 3.00

A Priori Scalar: 0.26

Adjusted Geodetic Coordinates

Pt ID Latitude Longitude Height

(m)

Height

Error

(m)

Fixed

NE N29°41'05.37214" W82°26'06.77104" 24.679 ? LLh

NW N29°41'05.61902" W82°26'14.56092" 28.765 0.002

SE N29°41'00.94168" W82°26'07.22740" 23.597 0.002

Adjusted GPS Observations

Observation

ID Observation

A-

posteriori

Error

Residual Standardized

Residual

NE --> NW

(PV3) Az. 272°04'45" 0.543 sec

0.160

sec 0.892

ΔHt. 4.086 m 0.002 m -0.001 m -1.725

Ellip

Dist. 209.577 m 0.001 m 0.000 m 0.657

NW --> SE

(PV1) Az. 126°08'42" 0.420 sec

-0.044

sec -0.267

ΔHt. -5.168 m 0.002 m -0.001 m -1.720

Ellip

Dist. 244.167 m 0.001 m 0.000 m -0.876

NE --> SE

(PV2) Az. 185°08'23" 1.199 sec

-0.327

sec -0.749

ΔHt. -1.082 m 0.002 m 0.001 m 1.705

Ellip

Dist. 136.968 m 0.001 m 0.000 m 0.833

Error Ellipse Components

Point ID Semi-major axis

(Meter)

Semi-minor axis

(Meter) Azimuth

NW 0.001 0.001 120°

SE 0.001 0.001 119°

SFe-NE SFe-NW

SFe-SE

GNVL

Figure 2: Test Site of a Base and Three GCPs

Figure 3: Test Reference Network

Base

NW NE

SE

SFe

15 km

Table 1: Reference Network Adjustment Report Summary

G1-m1TM Field Testing – Static Case

Test Procedures

Two G1-m1TM receivers were used along with a G1-

A311TM antenna in this test. Each receiver was running

on the SFe-NW check point concurrent with two Trimble

base receivers on the SFe-NE and SFe-SE stations which

allowed for acquiring two data sets for testing. Figure 4

shows the satellite visibility and DOPs as acquired by the

G1-m1TM system.

Data Set 1: In this case, the G1-m1TM AAC receiver was

setup on check point SFe-NW. 30 minutes of static data

was collected at 1-s interval, a total of 1972 epochs.

Data Set 2: Receiver G1-m1TM AAA was then setup on the

SFe-NW check point. It collected data for about 15

minutes at 1-s interval, a total of 760 epochs.

Test Results

HorusTM GNSIS post-mission processing software was

used to process the acquired data against the base

station data.

Short vs. Long Baselines: The acquired data was first

processed in static mode and the calculated position was

compared to the ‘known’ surveyed coordinates of the

occupied station (SFe-NW); table 2 summarizes the

results of the static processing of the two data sets.

As shown in Table 2, the results are consistent for the

short baselines between the systems and the SFe base (<

250 m). The horizontal error is in the cm-level, while the

vertical error has maxed to 3 cm.

The long baselines between the systems and the

Gainesville airport CORS station (GNVL) 15 km away yield

mixed results. While cm-level accuracy was not

achievable over the shorter duration of less than 15

minutes, the 30 minute session was, however, acceptable

with cm-level horizontal accuracy and 5 cm vertical.

Baseline DN [m] DE [m] Dh [m]

AAC Receiver

AAC11-SE 0.001 0.009 0.009

AAC11-NE 0.000 0.007 0.000

AAC11-GNVL 0.004 0.002 0.050

AAA Receiver

AAA06-SE 0.007 0.002 0.034

AAA06-NE 0.009 0.003 0.013

AAA06-GNVL 0.021 0.226 0.247

Static vs. Kinematic Data Processing: To evaluate the

behavior of the acquired data between epochs, the static

data sessions were processed one more time in kinematic

positioning mode, i.e. as if the receiver was moving.

As shown in Figure 5, the ground track of the system

position conforms to the results of the static processing.

Apart from a couple of epochs, almost 100% of the

positions were calculated based on carrier phase integer

ambiguity (green color) with a standard deviation and

RMSE in the cm level both horizontally and vertically; this

achievable for both data sets.

The acquired carrier phase ranges did not show any

particular pattern for this case, as shown in Figure 5. The

root-mean-squared error of the carrier phase residuals is

5 mm in this case, as expected.

Figure 4: Satellite Visibility and DOPs during the Test

Table 2: Summary Results of the Static Baselines (6 total)

G1-m1TM Field Testing – Static Case

The ground track and the residual errors, however,

showed some pattern in the longer baseline case, as

depicted in Figure 6. Almost 100% of the carrier phase

ambiguities were not resolved to integer value and a float

solution was used (amber color.)

The quasi-sinusoidal pattern is most likely due to the

atmospheric errors between the GNVL CORS and the SFe-

NW station (~ 15 km). The residual errors are

pronounced more in the up component, which indicates a

higher tropospheric effect. The other possible reason is

the relatively short session duration (~ 30 min.) over the

relatively long baseline length, where satellite-receiver

geometry was not allowed to change to de-correlate the

errors.

Conclusion

The results of this test show that cm-level horizontal and

vertical positional accuracy is achievable with the G1-

m1TM receiver in static mode over short baseline.

Although the collected data was relatively short in time

(only 15&30 minutes), it was adequate to show the

system potential.

On the other hand, cm-level positional accuracy over long

baseline was only achieved with longer data sessions;

sub-foot positional accuracy is always achievable. Longer

static sessions of the G1-m1TM system of two hours or

more is recommended for long baselines.

Figure 5: Ground Track and Residual Range Errors (Short Baseline)

Figure 6: Ground Track and Residual Range Errors (Long Baseline)

Contact us 2019 NW 21st Ln, Gainesville, FL 32605 Phone: (352) 226-9564

info@geomatics.US

Visit us on the web:

www.geomatics.us