DOP and Pseudorange Error Estimation in Urban Environments

for Mobile Android GNSS Applications

Przemyslaw Falkowski-Gilski

and Zbigniew Lubniewski

Faculty of Electronics, Telecommunications and Informatics, Gdansk University of Technology,

Narutowicza 11/12, Gdansk, Poland

Keywords: Mobile Devices, Outdoor Positioning, Smartphone-based Positioning, Wireless Positioning.

Abstract: Just a couple of years ago, GNSS (Global Navigation Satellite Systems) were available only for a narrow

group of users. Currently, with the outbreak of mobile devices, they are accessible to anyone and everywhere.

Urban navigation or searching for POIs (Points of Interest) had become an everyday activity. With the

availability of consumer electronics and wireless technologies, each user can obtain information considering

his or her location even in an unknown environment. Additionally, network operators and service providers

utilize this location-based information for monitoring and maintenance purposes. This paper is focused on a

study, considering the DOP (Dilution of Precision) and pseudorange error estimation in case of Android-

powered smartphones operating outdoors. It describes a measurement campaign, carried out in varying urban

environments, with two types of excursions (by car and bicycle), including two popular consumer devices

from different manufacturers. Based on this, respective conclusions and remarks are given. This work aims

to aid not only users, but also application developers as well as device manufacturers and retailers, when it

comes to providing precise and reliable products and services.

1 INTRODUCTION

Nowadays, there is a wide group of GNSS (Global

Navigation Satellite Systems) operating around the

world. They utilize artificial satellites, constantly

generating signals and transmitting them to the

surface of the Earth. Mobile devices, particularly

smartphones with a complex set of wireless modules,

enable to take advantage of a number of available

national and international systems, including: GPS

(USA), GLONASS (Russia), Galileo (Europe/EU),

BeiDou (China), NAVIC (India), and QZSS (Japan)

(Quan, Lau, Roberts and Meng, 2016; Teunissen and

Montenbruck, 2017). A brief description of GNSS,

including utilized frequency band, is described

in Table 1.

It seems that the variety of GNSS-based

applications is unlimited. Currently, they are utilized

in e.g. forwarding and logistics, aerial, road and sea

transport, personal motorized and pedestrian

navigation, searching for POI (Point of Interest), etc.

Particularly, users rely on them in urban

https://orcid.org/0000-0001-8920-6969

https://orcid.org/0000-0002-2821-1980

environments, especially at unknown locations.

It is worth mentioning that currently every mobile

device has an integrated circuit, responsible for

processing GNSS signals (Gilski and Stefański, 2015;

Chruszczyk, 2017).

Due to the popularity and widespread of portable

devices, it seemed interesting to investigate the

precision that modern smartphones can offer.

Particularly, what is the accuracy of satellite

positioning and navigation systems for urban

mobility applications. That is why this study,

concerning Android-powered terminals, was carried

out.

2 MOBILE CONSUMER

DEVICES

Due to the technological development, mobile

terminals have evolved into functionally-

sophisticated devices, such as smartphones.

Falkowski-Gilski, P. and Lubniewski, Z.

DOP and Pseudorange Error Estimation in Urban Environments for Mobile Android GNSS Applications.

DOI: 10.5220/0010719300003058

In Proceedings of the 17th International Conference on Web Information Systems and Technologies (WEBIST 2021), pages 629-638

ISBN: 978-989-758-536-4; ISSN: 2184-3252

629

The Android platform is currently the most popular

operating system worldwide.

Table 1: Modern GNSS satellite systems.

No. Name Frequency band

1 GPS 1.563–1.587 GHz (L1)

1.215–1.2396 GHz (L2)

1.164

–

1.189 GHz

(

)

2 GLONASS 1.593–1.610 GHz (G1)

1.237–1.254 GHz (G2)

1.189

–

1.214 GHz

(

)

3 Galileo 1.559–1.592 GHz (E1)

1.164–1.215 GHz (E5a/b)

1.260

–

1.300 GHz (E6)

4 BeiDou 1.561098 GHz (B1)

1.589742 GHz (B1-2)

1.20714 GHz (B2)

1.26852 GHz

(

)

5 NAVIC 1.17645 GHz(L5)

2.492028 GHz

(

)

6 QZSS 1.57542 GHz (L1-C/A)

1.57542 GHz (L1C)

1.2276 GHz (L2C)

1.17645 GHz (L5)

1.57542 GHz (L1-SAIF)

1.27875 GHz (LEX)

As handheld devices become more popular,

the role of an operating system grows significantly.

Current Android-powered devices are full of

integrated hardware, including IMUs (Inertial

Measurement Units) such as cameras, gyroscopes,

accelerometers, as well as various wireless interfaces,

containing a build-in cellular and GNSS receiver.

This OS enables third-party applications to make use

of these hardware features and provides a suitable

user environment (Gilski and Stefański, 2016).

2.1 Programming Android Terminals

As learning mobile programming becomes an

increasingly sought after and valued skill, primary,

secondary, as well as higher education institutions

aim at designing and developing courses, books and

related supplementary learning resources

(Hendikawati, Zahid and Arifudin, 2019a), along

with modern ICT (Information and Communication

Technologies) tools (Hendikawati, Zahid and

Arifudin, 2019b).

In (Prabowo, Rahmawati and Anggoro, 2019)

authors developed an Android-based application for

teaching junior high-school mathematics, integrated

with a popular social-media platform, namely

WhatsApp. Whereas, in (Hendikawati, Arifudin and

Zahid, 2018) authors developed a computer-assisted

application for statistical data analysis, also related to

education and learning purposes.

In (de Oliveira, de Oliveira, Ramalho and Viana,

2016) authors assessed the performance of mobile

messaging, under various research scenarios,

including different distributions of the Android OS,

mobile devices, as well as wireless network access

technologies.

Android itself, as an open environment, is a

member of the Linux family. The difference between

the Android and Linux kernel, runtime environment,

security and privacy risks, etc., is discussed in (Khan

and Shahzad, 2015). Whereas, the matter of mobile

security is further analyzed in (Ul Abideen, Ali Tariq,

Shah Talha Naqash and Qaseem, 2018).

As shown, the Android operating system,

and related consumer devices, are utilized for a

number of purposes, even in testing and monitoring

electrical components of machines. The matter of

such an IoT (Internet of Things) system, as an

element of the Industry 4.0 concept, is described

in (Sharmilah et al., 2019).

2.2 Positioning Accuracy

The positioning accuracy of a GNSS system may be

evaluated in a number of ways. The most popular one

is based on CNR (Carried-to-Noise-Ratio), usually

expressed in dBHz (Cisco, 2012). Another approach

is related to the number of observed and monitored

satellites, including the DOP (Dilution of Precision)

coefficient, related to the geometry of the

constellation, and its impact on precision (Langley,

1999). The DOP can be defined in a number of

variants:

 GDOP (Geometric DOP) – related to the

positioning accuracy in 4 dimensions

(space and time).

 HDOP (Horizontal DOP) – related to the

positioning accuracy in the horizontal plane.

 VDOP (Vertical DOP) – related to the

positioning accuracy in the vertical plane

(height).

 TDOP (Time DOP) – related to the accuracy of

time measurements.

 PDOP (Position DOP) – related to positioning

accuracy in 3 spatial dimensions.

Generally speaking, the lower the value,

the higher the reliability (confidence) of positioning

calculations. However, when DOP is close to or equal

to zero, obtaining a position is not possible, due to too

low signal level or severe interference. Whereas,

when it reaches a threshold of 20 and above,

QQSS 2021 - Special Session on Quality of Service and Quality of Experience in Systems and Services

630

the signal quality does not provide reliable measures

as well.

Before 2016, Android-powered devices, up to

Android Marshmallow 6.0, had limited access to

navigation and/or positioning data. At that time,

the API (Application Programming Interface)

enabled to access basic information concerning the

satellite’s azimuth, elevation, SNR (Signal-to-Noise

Radio), PVT (Position, Velocity and Time), status of

the chipset (active or inactive), latitude and longitude

data, estimated positioning accuracy (in meters),

as well as NMEA (National Marine Electronics

Association) data. This enabled a positioning

accuracy of a couple of meters (most often 2-3 m).

Since 2017, the newly introduced version of

Android, called Nougat (7.0), enabled to utilize raw

positioning data (GSA, 2017). From then developers

were able to use a set of dedicated additional classes

and related methods in order to design more precise

software, including information about the GNSS

clock (for pseudorange error calculation purposes),

how to decode incoming bits from available satellite

constellations, as well as the time, code and phase of

a particular carrier from a given satellite, not to

mention the Doppler shift.

Further improvements included integrating the

PPP (Precise Point Positioning) technique

(Laurichesse, Rouch, Marmet and Pascaud, 2017),

as well as sensor fusion, together with the mobile

device’s build-in light and pressure sensor,

IMU (Inertial Measurement Unit), accelerometer,

gyroscope, magnetometer, etc. Recent studies related

to the subject of mobile positioning are available

in (Specht, Szot, Dąbrowski and Specht, 2020;

Su, Jin and Jiao, 2020; Guo et al., 2020; An, Meng

and Jiang, 2020).

3 ABOUT THE STUDY

The study was carried out using two mobile devices,

particularly smartphones. They came from different

manufacturers, and are further labeled as Smartphone

1 and Smartphone 2.

The first one had a 8-core CPU (2.2 GHz),

3 GB of RAM, a 3000 mAh battery, and was powered

by Android Pie (9.0). The integrated GNSS module

was compatible with GPS, GLONASS, Galileo,

BeiDou, and QZSS.

The second device had also a 8-core CPU

(2.3 GHz), 4 GB of RAM, 4000 mAh battery, and was

GNSS module was compatible with GPS,

GLONASS, Galileo, and BeiDou.

All obtained data were recorded in the raw format,

and then processed using the GNSS Measurement

Tool as well as Matlab software. During both

measurement and processing, we utilized our custom

software, in order to obtain as much data as possible.

The measurement campaign included 2 types of

routes (square-shaped and straight line), 2 types of

communication means (bicycle – low speed, car –

high speed), and of course 2 smartphones.

4 RESULTS

In the first scenario, the route resembled a square,

as shown in Figure 1. Whereas in the second one,

the route resembled a straight line, as shown

in Figure 2. Both routes were evaluated with a bicycle

at an average speed of 8 km/h, and a car at average

speed of 35 km/h. This route went along 3 streets

(Twarda, Chwaszczyńska, and Okrąg) in the city of

Gdańsk. It varied in type of structure and its closest

neighborhood. Some part of it was surrounded by

buildings, and some of it was next to an open terrain.

Figure 1: Layout of the square-shaped route.

Figure 2: Layout of the straight line route.

To start with, obtained results will be described

taking into consideration the type of route (square-

shaped and straight line), type of mobility (bicycle

and car), and utilized smartphone (Smartphone 1 and

Smartphone 2). Next, obtained results will be

compared and discussed.

DOP and Pseudorange Error Estimation in Urban Environments for Mobile Android GNSS Applications

631

Figure 3: Pseudorange error – bike excursion along the square-shaped route with Smartphone 1.

4.1 Square-shaped Route: Bicycle

Excursion with Smartphone 1

The HDOP value, for the GPS constellation,

was close to approx. 1 in case of 7-10 observed

satellites. When the number of satellites decreased,

the HDOP value rose to approx. 2. In case of the

GLONASS system the number of observed satellites

dynamically changed between 5-8 satellites, whereas

HDOP ranged between 0.8-1.56. The clock drift did

fluctuate, with a maximum value of approx.

0.03 ppb/s.

The signal with the strongest CNR was observed

for satellite C27 from the BeiDou constellation.

This satellite, like C28 and C22, was observed only

in the first few seconds of measurement. The signal

strength from other BeiDou satellites was equal to

approx. 20-25 dBHz. However, those signals had a

big pseudorange error (from -6 to 4 m), and were not

correlated to the estimated frequency clock.

The signal strength itself was unstable.

The pseudorange error did exceed 50 m,

especially at the crossroad of Twarda and

Chwaszczyńska streets. To sum up, 95% of the

pseudarange error was less than 59.6 m. The WLS

(Weighted Least Squares) error was equal to 15.1 m

(see Figure 3). This route runs between buildings only

in some part. That is why the impact of multipath

propagation was not significant. Furthermore,

the terrain topology also had an impact on obtained

accuracy, as the height difference ranged between

0-4 m above sea level.

4.2 Square-shaped Route: Bicycle

Excursion with Smartphone 2

The absolute value of the clock drift was equal to

0.01 ppb/s. Most of the observed satellites had good

geometry. In case of GPS the HDOP value oscillated

around 0.9-1.2, whereas for GLONASS it reached

1-2.25.

The horizontal and vertical error, calculated based

on WLS, oscillated around 0.0-0.7 m. However,

this was not close to real values. According to

obtained results, the amplitude of height reached

100 m, which is a huge error, as the real value was

equal to approx. 4 m along the whole route

Obtained pseudorange errors were unconcise.

They changed from positive to negative (see

Figure 4). This clearly showed that positioning

accuracy was low. Higher Doppler shifts were

observed for GLONASS. It should be pointed out that

BeiDou and Galileo satellites were not monitored

during most of the time. Biases in the clock itself had

a significant impact as well.

QQSS 2021 - Special Session on Quality of Service and Quality of Experience in Systems and Services

632

Figure 4: Pseudorange error – bike excursion along the square-shaped route with Smartphone 2.

4.3 Square-shaped Route: Car

Excursion with Smartphone 1

The HDOP value for GPS was equal to approx. 1,

and reached 2 when the number of satellites shrank

to 4. In case of GLONASS, this parameter oscillated

from 1 to even 5. During approx. half of the time, the

clock was not concise, resulting in a discontinuous

time of satellite observation, especially in case of

BeiDou and Galileo. Not surprising, the strongest

signal was observed when driving in open terrain (the

number of observed satellites was also higher).

During the measurement campaign, the clock was not

stable. Moreover, a frequency drift was observed,

equal to 0.18 ppb/s, with a stable clock bias.

The pseudorange error for GPS and GLONASS

oscillated from -50 to 50 m (see Figure 5).

When driving in open space, additional Galileo and

BeiDou signals were observed. The highest

pseudorange error was observed when the car was in

the so-called urban canyon.

The WLS estimation was quite precise, especially

in the second part of the drive test. At first even

deviations of few meters were observed. During the

first 52 s the vertical position was calculated with

huge error (exceeding 5 m). Whereas, for the next

40 s, the accuracy was noticeably higher. However,

at the end it reached 50 m. When utilizing WLS on

raw data, we obtained more accurate results.

The average vertical position was approx. at 0 m,

whereas 50% of obtained samples resulted in 6.3 m

and less.

4.4 Square-shaped Route: Car

Excursion with Smartphone 2

In case of GPS the HDOP value oscillated around

approx. 0.9-1.5, whereas in case of GLONASS it did

not exceed 1.6. The clock was continuous, with a drift

of -0.06 ppb/s. The clock shift was getting higher in a

linear scale.

The pseudorange error, for most cases (95%)

did not exceed 28.7 m. When the car drove among

buildings, this error reached to approx. 100 m

(see Figure 6). The highest Doppler shifts were also

observed for G07 and G08 (30-50 s and 90-100 s),

where the observed multipath effect was the

strongest.

Results obtained using WLS were quite good.

The horizontal error was less than 9.7 m for 50% of

obtained samples, and less than 20.8 m for 95% of

samples. The vertical error was less than 23.4 m for

50% of samples, and less than 44.6 m for 95% of

samples.

DOP and Pseudorange Error Estimation in Urban Environments for Mobile Android GNSS Applications

633

Figure 5: Pseudorange error – car excursion along the square-shaped route with Smartphone 1.

Figure 6: Pseudorange error – car excursion along the square-shaped route with Smartphone 2.

4.5 Straight Line Route – Bicycle

Excursion with Smartphone 1

The HDOP value in case of GLONASS oscillated

from 1.2 to 3. The signals from other satellite systems

were stable. In some part of the measurement time,

reception of Galileo and BeiDou was not possible,

due to signal loss.

At first, the frequency drift rose with 0.45 ppb/s,

then shrank with 0.15 ppb/s. Between 80-140 s,

the receiver did not monitor satellites in a continuous

way. That is why during this time the clock drift was

not calculated. In other parts of the measurement

time, the drift changed with -0.24 ppb/s

(when moving from building surroundings to open

space). The clock drift, calculated between 86-110 s,

was equal to 593 ppm.

The biggest observed pseudorange error was

equal to 150 m (see Figure 7). When surrounded by

high buildings, we were not able to determine the

QQSS 2021 - Special Session on Quality of Service and Quality of Experience in Systems and Services

634

Figure 7: Pseudorange error – bike excursion along the straight line route with Smartphone 1.

Figure 8: Pseudorange error – bike excursion along the straight line route with Smartphone 2.

error. A linear decrease in error may be observed,

when the device reached open space. Position

calculations using WLS provided an accurate

estimation. Between 80-140 s, clock biases were

encountered. The horizontal error for 50% of results

was less than 164.5 m, and less than 235.1 m for 95%

of data. In case of vertical error, it was equal to less

than 28.4 m for 50% of data, and less than 215 m for

95% of data.

4.6 Straight Line Route – Bicycle

Excursion with Smartphone 2

In case of the HDOP values for GPS, they oscillated

around approx. 1.2, whereas GLONASS received

approx. 2. This mobile device enabled stable

reception for GPS and GLONASS constellations.

The CNR value did not change much, equal too

approx. 38-45 dBHz over time.

DOP and Pseudorange Error Estimation in Urban Environments for Mobile Android GNSS Applications

635

Figure 9: Pseudorange error – car excursion along the straight line route with Smartphone 1.

The frequency drift was equal to 0.04 ppb/s.

The clock bias raised in a linear scale, up to 12 ms.

The pseudorange error in case of 95% of data was less

than 35.7 m. The pseudorange error oscillated around

0 m/s, and was getting higher for satellites on low

elevation levels (see Figure 8).

In case of the WLS, the horizontal error for 50%

of samples was less than 9.3 m, and less than 18.2 m

for 95% of samples. The vertical error was equal to

16.1 m for 50% of samples, and 40.5 m for 90%

of samples. The estimated height was equal to above

60 m, which was not close to real conditions, in which

the height ranged for approx. 2 m. In case of raw data,

the estimated height was around 0 m, with 40 m

in just some cases.

4.7 Straight Line Route – Car

Excursion with Smartphone 1

In case of BeiDou, the C19 satellite exceeded the

reference signal level. However, the signal from this

satellite was only observed for a limited time period.

The HDOP value ranges from 1.2 to 3. For GPS it

indicated 7 to 11 objects, whereas in case of

GLONASS, the number of monitored satellites

ranged from 3 to 9.

During the evaluation, the clock was not stable, the

maximum frequency drift was equal to 11.94 ppb.

The clock drift was equal to 512 ppm. Due to clock

biases, the pseudorange error in case of 95% of

samples was less than 20.7 m. For Galileo, the error

raised up to 100 m (see Figure 9). Thanks to WLS,

the horizontal error for 95% of samples was less than

7.7 m, whereas for 50% it was less than 0.3 m.

4.8 Straight Line Route – Car

Excursion with Smartphone 2

In case of GPS, those satellites had good geometry.

HDOP during 63 s achieved a value of less than 1.

Whereas, the number of observed satellites was equal

to 10. The received signal strength level was stable,

especially when examining GPS satellites at a height

of above 30 degrees, i.e. G12.

During the whole time, the clock was stable,

for the first 27 s the drift was relatively stable,

increasing by 0.04 ppb/s. The clock bias was

increasing linearly, reaching 4.5 ms at the end.

The pseudorange error for 95% of samples was

less than 13.5 m (see Figure 10). This is mostly due

to the fact that measurements were carried out in an

open space. The Doppler shift error oscillated from

0 up to 3 m/s.

In case of WLS, the horizontal error for 95%

of samples did not exceed 6.8 m, and was less than

3.3 m for 50% of samples. Whereas the vertical error,

in case of 50% of samples, was less than 4.5 m,

whereas for 95% of data it was less than 13.5 m.

According to obtained results, for the first 40 s,

the receiver (mobile device) was traveling on flat

terrain, then the height decreased. However, this was

not correct, as the height itself did not change.

QQSS 2021 - Special Session on Quality of Service and Quality of Experience in Systems and Services

636

Figure 10: Pseudorange error – car excursion along the straight line route with Smartphone 2.

5 SUMMARY

This work describes results of a study, focused on the

DOP and pseudorange precision of GNSS systems

using smartphones, their quality, reliability, related to

signal reception. The tested devices, coming from

2 different manufacturers, were all Android-powered

devices. The measurement campaign itself was

carried out in different conditions, including

surrounding buildings, terrain topology, and urban

fabric.

5.1 GNSS Signal Reception

The first smartphone seldom provided a signal level

above the referenced CNR. Whereas the second one

received at least over 10 such signals (above

reference level), from which only 1 came from a

constellation other than GPS or GLONASS.

The CNR value itself was also different. In case

of Smartphone 1 this parameter dynamically changed

(fluctuated) over time, even in a few second interval.

On the other hand, Smartphone 2 proved to be more

stable, as recorded CNR values did not changed that

rapidly.

5.2 Time Calculations and Precision

One must note that maintaining continuity of clock

(time) measurements is of great importance. In case

of the second device, no discontinuous measurements

were observed. In most cases, the frequency drift was

equal to approx. 0.058 ppb/s. Whereas in case of

Smartphone 1, during 26.6% of time,

the measurements were discontinuous. In this case,

most often the average frequency drift was equal to

approx. 0.252 ppb/s. Furthermore in one case, it was

not possible to calculate the drift due to instability.

With the utilization of the WLS algorithm, it was

possible to calculate the pseudorange value. In case

of Smartphone 1, the average value was equal to

31.82 m. Whereas for Smartphone 2, it was equal to

32.5 m. As shown, this difference was slight.

5.3 Overall Remarks

According to the study, it should be pointed out that

numerous factors and conditions influence the final

accuracy and precision of satellite positioning,

especially in urbanized areas, when using different

means of transport. The navigation performance

depends on basic characteristics of the smartphone,

both communication (e.g. antenna features, supported

wireless systems, etc.) and data processing

(CPU, RAM, storage, etc.). In such a context, it seems

interesting to evaluate and compare in practice a

broader range of mobile devices in future studies.

As noticed, more and more powerful consumer

electronics are released on the market every year.

On the other hand, user preferences and expectations

do change from time to time (Finley et al., 2017;

Falkowski-Gilski, 2020; Falkowski-Gilski and Uhl,

DOP and Pseudorange Error Estimation in Urban Environments for Mobile Android GNSS Applications

637

2020). Future studies should include consumer

devices with diverse integrated GNSS modules,

different distribution of the Android operating

system, as well as deployment scenarios.

REFERENCES

An, X., Meng, X., Jiang, W. (2020). Multi-constellation

GNSS precise point positioning with multifrequency

raw observations and dual-frequency observations of

ionospheric-free linear combination. Satellite

Navigation, 1, 7.

Chruszczyk, Ł. (2017). Statistical analysis of indoor RSSI

read-outs for 433 MHz, 868 MHz, 2.4 GHz and 5 GHz

ISM bands. International Journal of Electronics and

Telecommunications, 63(1), 33-38.

Cisco (2012). Digital Transmission: Carrier-to-Noise

Ratio, Signal-to-Noise Ratio, and Modulation Error

Ratio.

Falkowski-Gilski, P. (2020). On the consumption of

multimedia content using mobile devices: A year to

year user case study. Archives of Acoustics, 45(2),

321-328.

Falkowski-Gilski, P., Uhl, T. (2020). Current trends in

consumption of multimedia content using online

streaming platforms: A user-centric survey. Computer

Science Review, 37, 100268.

Finley, B., Boz, E., Kilkki, K., Manner, J., Oulasvirta, A.,

Hämmäinen, H. (2017). Does network quality matter?

A field study of mobile user satisfaction. Pervasive and

Mobile Computing, 39, 80-99.

Gilski, P., Stefański, J. (2015). Survey of radio navigation

systems. International Journal of Electronics and

Telecommunications, 61(1), 43-48.

Gilski, P., Stefański, J. (2016). Android OS: A review,

TEM Journal, 4(1), 116-120.

GSA. (2017). Using GNSS raw measurements on Android

devices.

Guo, L., Wang, F., Sang, J., Lin, X., Gong, X., Zhang, W.

(2020). Characteristics analysis of raw multi-GNSS

measurement from Xiaomi Mi 8 and positioning

performance improvement with L5/E5 frequency in an

urban environment. Remote Sensing, 12, 744.

Hendikawati, P., Arifudin, R., Zahid, M. Z. (2018).

Development of computer-assisted instruction

application for statistical data analysis android platform

as learning resource. Journal of Physics: Conference

Series, 983, 012057.

Hendikawati, P., Zahid, M. Z., Arifudin, R. (2019a).

The development of textbooks for mobile application

programming. Journal of Physics: Conference Series,

1321, 032019.

Hendikawati, P., Zahid, M. Z., Arifudin, R. (2019b)

Android-based computer assisted instruction

development as a learning resource for supporting self-

regulated learning. International Journal of Instruction,

12(3), 389-404.

Khan, J., Shahzad, S. (2015). Android architecture and

related security risks. Asian Journal of Technology and

Management Research, 5(2), 14-18.

Langley R. B. (1999). Dilution of precision. GPS World,

5, 52-59.

Laurichesse, D., Rouch, C., Marmet, F. X., Pascaud, M.

(2017). Smartphone applications for precise point

positioning. In ION GNSS+ 2017, 30th International

Technical Meeting of the Satellite Division of The

Institute of Navigation, 171-187.

de Oliveira, R. C., de Oliveira, H. M., Ramalho, R. A.,

Viana, L. P. S. (2016). Performance assessment of

WhatsApp and IMO on Android operating system

(Lollipop and KitKat) during VoIP calls using 3G or

WiFi. Global Journal of Computer Science and

Technology, 16(1), 1-8.

Prabowo, A., Rahmawati, U., Anggoro, R. P. (2019).

Android-based teaching material for statistics

integrated with social media WhatsApp. International

Journal on Emerging Mathematics Education, 3(1),

93-104.

Quan, Y., Lau, L., Roberts, G.W., Meng, X. (2016).

Measurement signal quality assessment on all available

and new signals of multi-GNSS (GPS, GLONASS,

Galileo, BDS, and QZSS) with real data. Journal of

Navigation, 69, 313-334.

Sharmilah, T., Yahya, M., Fuzi, K. M., Pauzi, M.,

Shirlynda, Z., Faris, N. A. (2019). Development of

temperature monitoring towards Industry 4.0.

In APCOM’19, 25th International Conference on

Applied Physics of Condensed Matter, 020034.

Specht, C., Szot, T., Dąbrowski, P., Specht, M. (2020).

Testing GNSS receiver accuracy in Samsung Galaxy

series mobile phones at a sports stadium. Measurement

Science and Technology, 31(6), 064006.

Su, K., Jin, S., Jiao, G. (2020). Assessment of multi-

frequency global navigation satellite system precise

point positioning models using GPS, BeiDou,

GLONASS, Galileo and QZSS. Measurement Science

and Technology, 31(6), 064008.

Teunissen, P. J. G., Montenbruck, O. (Eds.) (2017).

Springer Handbook of Global Navigation Satellite

Systems, Springer. Berlin, 1

edition.

Ul Abideen, Z., Ali Tariq, H., Shah Talha Naqash, S. H.,

Qaseem, U. (2018). Android apps management system

to ensure mobile security. In ICKII’18, 1

IEEE

International Conference on Knowledge Innovation

and Invention, 78-81.

QQSS 2021 - Special Session on Quality of Service and Quality of Experience in Systems and Services

638