See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/339536906

Low Cost GNSS Receiver RTK Performance in Forest Environment

Conference Paper · February 2020

DOI: 10.23919/URSIRCRS49211.2020.9113621

CITATIONS READS
28 800

6 authors, including:

Somnath Mahato Gopal Shaw

India Meteorological Department University of Burdwan
67 PUBLICATIONS 364 CITATIONS 1 PUBLICATION 28 CITATIONS

SEE PROFILE SEE PROFILE

Atanu Santra Sukabya Dan

Collins Aerospace University of Burdwan
53 PUBLICATIONS 341 CITATIONS 49 PUBLICATIONS 302 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Somnath Mahato on 20 September 2022.

The user has requested enhancement of the downloaded file.

 URSI RCRS 2020, IIT (BHU), Varanasi, India, 12 - 14 February, 2020

Low Cost GNSS Receiver RTK Performance in Forest Environment

Somnath Mahato(1,2), Gopal Shaw(1), Atanu Santra(1,2), Sukabya Dan(1), Surajit Kundu(2) and Anindya Bose(1)
(1) GNSS Laboratory, The University of Burdwan, Burdwan - 713 104, India
(2) National Institute of Technology Sikkim, Ravangla - 737 139, India
Email: [email protected]

Abstract surveying. Precise positioning using (RTK) is a preferred

technique in surveying purposes.
Requirement of accuracy and reliability of position
solution obtained from satellite-Based navigation system GNSS RTK position solution is calculated using carrier
provided by the Global Navigation Satellite System phase-based measurements providing better precision than
(GNSS) depends on the application. In some cases, meter its code-based counterpart [1]; RTK solution is classified
level precision is required while cm or mm level accuracy into two types: Fix and Float [2]. Under open sky condition,
is required for other precision applications. Surveying is RTK performs well; an efficiently designed RTK setup
one of the important applications of GNSS. Surveying of provides centimeter level of solution accuracy [3, 4]. But
forest is a challenge for survey engineers in determining within the forest canopy, the satellite signals are
accurate position, because satellite signals frequently fade constrained by leaves and stems of trees. Also, multipath
by stems and leaves of the tress. Consequently, the signals reach the GNSS antenna being reflected from the
observables become noisy that affects quality of position trees and branches. The observables become noisy, and
solution. Real Time Kinematic (RTK) is a popular GNSS therefore integer ambiguity resolution faces difficulty to
positioning method used in case of various precision estimate the fixed integer value of number of carrier cycles.
applications. High cost receivers may provide better RTK Therefore, obtaining better position solution under forest
solution those are not affordable for many surveyors. Low canopy is a point of interest for the surveyors.
cost, single frequency, multi-GNSS enabled receiver are
available in market those can provide raw data and can be Centimeter level solution accuracy has been achieved in
used for RTK. In this paper an effort is presented to assess forest area using Field Mate systems on GPS only
the precision and reliability of position solution obtained constellation [5]. GPS+GLONASS RTK under forest
from the low-cost receivers within Indian forest environment has been reported in [6]. Costly, survey-grade
environments. In the first case study over a short baseline, receivers used for RTK may not be affordable for many
position solution precision provided by a low-cost Rover users, especially for the countries like India where scope
operating in forest is studied, and meter level of precision for utilizing GNSS is increasing. Usability of low-cost,
is achieved in GPS+GLONASS+Galileo mode. In second compact, single frequency multi-GNSS receivers in such
case study for long baseline RTK operation, reliable regions would be an attractive alternative. Over the last
solution is obtained in GPS+Galileo+QZSS mode from a couple of years, many receiver manufacturers have
low-cost Rover operating within forest environment. So, introduced compact, power and cost effective GNSS
reliable or precise position solution can be achieved using receivers of price around US$100, which can deliver raw
low cost receivers and suitable multi-GNSS combination and NMEA data output. The raw data can be used through
under forest canopy. appropriate software for RTK-based GNSS solution.
RTKLib is such an open source program package which
1. Introduction can perform RTK with the compact GNSS receivers [7].
RTK performance varies according to the Base-Rover
Global Navigation satellite system (GNSS) is increasingly distance (baseline length) because local atmospheric
being popular to obtain accurate and reliable position conditions and set of visible satellites may vary at the Base
solution using satellite signal. In the current multi-GNSS and the Rover locations over a long distance [8].
environment, the global systems are GPS (US DoD),
GLONASS (Russian), Galileo (European Union), BeiDou Asia-Oceania region is a good place for receiving multi-
(China) and the regional navigation systems are QZSS GNSS signal offering extra advantage for the surveyors of
(Japan) and NavIC (India). The multi-GNSS with large this region. GPS, GLONASS and Galileo are the prominent
Signal in Space (SiS) scenario is supporting enhancement global systems those are used for research in RTK
of quality of position solution and redundancy of satellite operations. GPS and Galileo are Code division Multiple
signals. GNSSs are currently widely used in myriads of Accessed (CDMA) signals, GLONASS signal design is
application; one of the most important applicable area is based on Frequency Division Multiple Access (FDMA)

Authorized licensed use limited to: COLORADO STATE UNIVERSITY. Downloaded on June 12,2020 at 10:36:39 UTC from IEEE Xplore. Restrictions apply.
 while QZSS is based on CDMA signal and RTK operation
is implemented on these constellations. In case of Multi-
GNSS operation, the inter system biases are to be taken
care of [9- 11]. In this paper an effort has been made to
study the precision of position solution of low cost, small
factor Rover receivers operating in forest environment for
(i) GPS+GLONASS+Galileo mode using RTK at short
baseline and (ii) performance of different multi-GNSS
combinations using RTK over long baseline. Here, the
combinations are (GPS+GLONASS+Galileo+QZSS; three
CDMA and one FDMA) and (GPS+Galileo+QZSS; all
three are CDMA). It may be noted that, use of single
constellation is restricted by the tree-canopy shadowing as
in such cases, seamless position solution gets interrupted
due to the signal blockage by the canopy. Therefore, in a
Multi-GNSS environment, a suitable option is the use of
more than one constellation together and this manuscript
compares the performances of various multi-GNSS
combinations for small and long baseline RTK.

To carry out these proposed works, two forest areas have

been chosen at different distances from the Base. A Figure 1. The locations of the Case studies and Antenna
permanent GNSS reference station consisting of a Leica of the Permanent reference station with NTRIP facility
GR50 receiver and Leica AR 25 antenna with Network at The University of Burdwan, India (inset).
Transport of Radio Technical Commission for Maritime
(RTCM) messages over Internet Protocol (NTRIP) caster connected with a computer through USB cable.
facility, situated at the rooftop of Department of Physics,
The University of Burdwan, Burdwan, India, RTKLib is an open source software that can use this raw
(23.25451923611°N, 87.846737522°E, -11.402 m) as data for RTK solution. RTKNAVI utility of RTKlib
shown in Figure 1, has been used as the static Base. Fixed program package, running on the computer connected to
coordinate of this reference station antenna was found Rover receiver, can use the raw data from the Rover
using PPP services provided by AUSPOS [12]. The case receiver over USB, and the Base-transmitted RTCM3
studies are presented in the next section. MSM7 messages to provide RTK position solution. Here
the Rover antenna (along with the receiver) is placed on a
2. Case I: Short Baseline RTK in forest tripod under forest canopy. The Ublox Rover receiver is
operated in GPS+GLONASS+Galileo hybrid multi-
constellation mode for 35 minutes on 21st May, 2019.
2.1 Introduction: Performance of RTK positioning is discussed in the next
subsection.
In this section, results of an experiment on RTK
performance of low-cost receivers at short baseline is
discussed. Location of study area is Ramnabagan Wildlife
Sanctuary, situated in Burdwan, West Bengal, India near to
The University of Burdwan campus. This area is located
around (23.522938° N, 87.493283° E). The highest height
of the trees is 22 meters and the average tree height is 15
meters in this forest those are measured using Theodolite.
A low-cost receiver with the antenna mounted at a fixed
location within the forest. The distance between Base and
Rover antennas is around 600 meters. The detail
experimental setup is described in the subsequent section.

2.2 Experimental setup:

The experimental arrangement for the study is shown in

Figure 2. This Base transmits RTCM3 Multiple Signal
Messages version 7 (MSM7) messages over internet [13]. Figure 2. Experimental setup for low-cost RTK
Ublox M8T low cost receiver with a multi-GNSS low-cost operation. The RTCM messages from the Base (NTRIP
antenna is used here as the Rover within the forest canopy. Caster) is sent to the Rover operating within forest
It provides raw data output in GPS, GLONASS, Galileo, environment over internet. The Baseline distance is
BeiDou and QZSS mode (maximum 32 Channels) when around 600m.

Authorized licensed use limited to: COLORADO STATE UNIVERSITY. Downloaded on June 12,2020 at 10:36:39 UTC from IEEE Xplore. Restrictions apply.
 2.3 Result and discussions: 2.2 above. Here the distance between Base and Rover
receiver is around 60 Kms. Two different multi-GNSS
For the position solutions provided by the Rover, 2D combinations have been considered here-one is
Precision parameters 2DRMS, CEP and 3D parameters GPS+Galileo+QZSS and other is GPS+GLONASS+
SEP and MRSE are calculated using the formulae depicted Galileo+QZSS. Data is collected for 30 min duration on 6
in [14]. The precision parameter values are shown in Table June, 2019 in each mode. The position solution obtained
1. It may be observed that, horizontal precision is 3.168 m from the Rover are analyzed for precision and the
and 3D precision is in meter level. Maximum variation in performances of the low-cost Rover in different multi-
altitude and longitude is bit higher than that of latitude. GNSS combination and is described in the following sub-
section.
Therefore, for short baseline RTK operation, with hybrid
GPS+GLONASS+ Galileo operation, these low-cost 3.2 Result and discussion:
GNSS receivers may provide meter level of precision The precision parameters are shown in Table 2 and the
within forest. The next effort is made to study the reliability scatter plot of solutions are shown in Figure 3. Maximum
of the RTK solution in case of long baseline under forest horizontal spread is observed in GPS+GLONASS+
canopy in different multi-GNSS combinations. Galileo+QZSS mode, as shown in Figure 3. From Table 2,
it is observed that 0.44% and 0.94% of Fixed RTK solution
Table 1. Precision parameters of RTK solution obtained are obtained in GPS+Galileo+QZSS and
from U-Blox M8T receiver in GPS+GLONASS+Galileo GPS+GLONASS+Galileo+ QZSS modes respectively. It
mode using RTKNAVI; Location; Raman Bagan is witnessed that 2DRMS, CEP, SEP, MRSE are better in
wildlife sanctuary, Burdwan, India (21th May 2019). GPS+Galileo+QZSS combination than those for
GPS+GLO+Galileo+QZSS mode. Maximum variation in
Precision parameter Values (m)
altitude changes sinificantly once GLONASS is added with
Maximum variation in latitude. 3.149
GPS+Galileo+QZSS combination. It is to be noted that,
Maximum variation in longitude 4.510
average latitude, longitude values are almost equal in all
Maximum variation in altitude. 4.400 modes but altitude values varies, as shown in Table 2. Here,
2DRMS (m) 3.168 reliable altitude should be nearer to 21.330 m which is the
CEP (m) 1.292 predetermined height of Rover antenna, previously
SEP (m) 1.716 measured using JAVAD Triumph LS Survey Grade
MRSE (m) 1.961 receiver in RTK mode.

3. Case II: Long baseline RTK under forest

canopy:
3.1 Introduction:

In this effort, effectiveness of various GNSS combinations

in long baseline RTK operation for position solution under
forest canopy is studied. Location of study area is Shibpur
forest of Durgapur, West Bengal, India (Figure 1) located
around (23°36'22"N, 87°25'42"E). Average tree height is
15 ± 20 meters with an average distance between two trees
of 4-5 meters. The forest is made up of predominant Figure 3. Horizontal scatter plot of solutions obtained
presence of Sal trees with other sub-tropical tree verities. from Ublox M8T Rover receiver under forest canopy
The experimental setup is like that depicted in subsection (Case 2)

Table 2. Precision of position solution obtained from RTK operation in different Multi GNSS combination at long baseline
( ~ 60 kms ) in forest area of Shibpur forest, Durgapur, India
RTK Maximum Variation (m) Avg. Avg. Avg.
Solution 2DRMS CEP SEP MRSE Lat Long Alt
Mode Lat Long Alt
(%) (m) (m) (m) (m) (deg) (deg) (m)
(m) (m) (m)
GPS+GAL+ Fix
3.393 4.112 3.070 5.137 2.114 2.570 2.941 23.5938 87.423 22.038
QZSS (0.44)
GPS+GAL+ Float
8.930 13.313 33.048 3.096 1.174 3.172 4.465 23.5938 87.423 22.718
QZSS (99.56)
GPS+GAL+ Fix
34.715 13.980 57.883 28.283 11.286 22.346 28.754 23.5939 87.424 26.185
GLO+QZSS (0.94)
GPS+GAL+ Float
52.836 22.490 93.148 26.263 10.523 21.477 27.875 23.5939 87.423 29.845
GLO+QZSS (99.06)

Authorized licensed use limited to: COLORADO STATE UNIVERSITY. Downloaded on June 12,2020 at 10:36:39 UTC from IEEE Xplore. Restrictions apply.
 From Table 2, it is observed that the obtained average [4] 6 0DKDWR $ 6DQWUD 6 'DQ $ %RVH ³/RZ-cost
height 22.0381 m in GPS+Galileo+QZSS (RTK Fix) is *166 PRGXOHV IRU 3UHFLVH 3RVLWLRQLQJ´ IEEE
closer to the predetermine height value. Figure 3 points International Conference on Range Technology, Chadipur,
towards the better solution in this combination. It may be India, 15-17 February 2019.
concluded that accuracy of position solution is more
reliable in GPS+Galileo+QZSS with less than 1% µ)L[¶ [5@ 0 %DNXáD 6 2V]F]DN DQG 5 3HOF-Mieczkowska,
RTK solution obtained in case of long baseline RTK. ³3HUIRUPDQFH RI 57. SRVLWLRQLQJ LQ IRUHVW FRQGLWLRQV
3UHFLVLRQLVEHWWHULQµ)ORDW¶VROXWLRQ, but accuracy is better &DVHVWXG\´Journal of Surveying Engineering, vol. 135,
LQµ)L[¶VROXWLRQ no. 3, pp.125-130, 2009. DOI: 10.1061/(ASCE)0733-
9453(2009)135:3(125)
4. Concluding remarks
[6] H. Andreas, H. Z. Abidin, D. A. Sarsito, and D.
This paper presents the initial results of multi-GNSS RTK 3UDGLSWD ³6WXG\ WKH FDSDELOLWLHV RI 57. 0XOWL *166
operation for two different baseline distances with low cost XQGHUIRUHVWFDQRS\LQUHJLRQVRI,QGRQHVLD´In E3S Web
receiver in forest environment where the GNSS signals are of Conferences, Vol. 94, pp. 1-6, 2019. DOI:
degraded due to shading and multipath. It is observed that, 10.1051/e3sconf/20199401021
meter level precision has been achieved in short distance
RTK. For Long-baseline RTK in forest environment, [7@77DNDVXDQG$<DVXGD³'HYHORSPHQWRIWKHORZ-
GPS+Galileo+QZSS mode performs better cost RTK-GPS receiver with an open source program
GPS+Galileo+GLONASS+QZSS mode. This may be SDFNDJH 57./,%´ In International symposium on
because of intersystem biases of heterogeneous receiver. GPS/GNSS, pp. 4-6. International Convention Center Jeju
The work presented here are preliminary and the studies Korea, 2009.
have been made under forest environment under static
condition with short and long baseline distance between the [8] T. Takasu, and A .Yasuda, ³.DOPDQ-filter-Based
Base and the Rover. But the results present the basis of integer ambiguity resolution strategy for long-Baseline
using GNSS RTK in Indian forest environments. Future 57. ZLWK LRQRVSKHUH DQG WURSRVSKHUH HVWLPDWLRQ´
studies would be extended to explore the performances of In Proceedings of the ION GNSS, pp. 161-171, 2010.
different cost-effective receivers and antennas used in the
Rovers within real-life operational conditions for varying [9] %. &KRL .0 5RK 6 - /HH ³$QDO\VLV RI WKH
distances. Multi-GNSS RTK would be explored further by combined positioning accuracy using GPS and GLONASS
adding BeiDou constellation in future. 1DYLJDWLRQVDWHOOLWHV´Journal of Positioning, Navigation
and Timing, vol.2 no.2, pp.131-137, 2013
5. Acknowledgments
[10@ $+ 'RGVRQ 7 0RRUH % ) %DNHU ³+\EULG
The authors acknowledge Defence Research and *36*/21$66´GPS Solution, vol. 3, no. 1, pp 32-41,
Development Organization (DRDO), New Delhi (Project 1999
Code: ERIP/ER/DG-MSS/990516601/M/01/1658) and All
India Council for Technical Education (AICTE), New [11@33U]HVWU]HOVNL0%DNXOD5*DODV³7KHLQWHJUDWHG
Delhi (Project Code: File N0. 8-10/ RFID/RPS/POLICY- use of GPS/GLONASS observations in network code
1/2016-17) for the financial support. GLIIHUHQWLDO SSRVLWLRLQJ´ GPS Solution, vol.21, no.2, pp
627-638, 2017
6. References
[12] R. Ghoddousi-)DUG DQG 3 'DUH ³2QOLQH *36
[1] Z. Xiuqiang, Z. Xiumei and C. YDQ³,PSOHPHQWDWLRQ
processing services: an initial study´GPS Solution, vol. 10,
RIFDUULHUSKDVHPHDVXUHPHQWVLQ*36VRIWZDUHUHFHLYHUV´
no. 1, pp.12-20. DOI: 10.1007/s10291-005-0147-5
2013 International Conference on Computational
Problem-solving (ICCP), Jiuzhai Valley, China, pp. 338-
[13] An RTCM 3 message cheat sheet, https://www.use-
341, October 26-28, 2013.
snip.com/kb/knowledge-Base/an-rtcm-message-cheat-
[2] 5%/DQJOH\³5TK GPS´GPS World, no. 9, pp. 70- sheet/, accessed on 13/10/2018.
76, 1998.
[14] A. Santra60DKDWR6'DQDQG$%RVH³Precision
[3] S. Mahato, A. Santra, S. Dan, P. Rakshit, P. Banerjee of Satellite Based Navigation Position Solution: a Review
DQG$%RVH³3UHOLPLQDU\5HVXOWVRQWKH3HUIRUPDQFHRI using NavIC Data´ International Conference In Recent
Cost-effectLYH *166 5HFHLYHUV IRU 57.´ URSI Asia- Trends on Electronics & Computer Science, Silchar, India,
Pacific Radio Science Conference , New Delhi, India, 18-19 March 2019.
2019. DOI: 10.23919/URSIAP-RASC.2019.8738736

Authorized licensed use limited to: COLORADO STATE UNIVERSITY. Downloaded on June 12,2020 at 10:36:39 UTC from IEEE Xplore. Restrictions apply.
View publication stats