Drive Test-based Correlation Assessment of QoS Parameters

for Exemplary Measurements Scenario in Suburban Environment

Dariusz Zmysłowski

and Jan M. Kelner

Institute of Communications Systems, Faculty of Electronics, Military University of Technology, Warsaw, Poland

Keywords: Quality of Service (QoS), Quality of Experience (QoE), Video Mean Opinion Score (VMOS), Signal-to-

Interference-plus-Noise Ratio (SINR), Reference Signal Received Power (RSRP), Drive Test, Pearson

Correlation Coefficient.

Abstract: The development of mobile networks is directly related to developing the telecommunications services market.

Introducing new services requires adjusting methods of quality of service (QoS) assessment and management.

QoS control measurements in mobile networks are a standard practice carried out by specialized companies

for mobile network operators and regulators of the telecommunications market. In this paper, we analyze a

selected measurement scenario for a suburban environment in which QoS was evaluated for video

transmission from the YouTube service. The Systemics-PAB Group company carried out the measurements

for the Polish regulator, the Office of Electronic Communications. The obtained measurement results are the

basis for the correlation analysis between the QoS parameters. The analysis results show a strong relationship

between the selected parameters, which can be used in modeling and simulation studies.

1 INTRODUCTION

The first mobile networks only provided voice call

services. With the development of the next

generations of mobile networks, new services have

been made available to users, including short message

service (SMS), multimedia messaging service

(MMS), or broadly understood packet data

transmission (Razeghi, 2007; Lloyd-Evans, 2002).

Offering new, diverse telecommunications services in

subsequent generations of cellular networks forced

the development, implementation, and improvement

of quality management systems for the provided

services (Oodan et al., 2002). In the modern third

(3G), fourth (4G), and fifth generation (5G) mobile

networks, the concept of quality is usually analyzed

in three aspects, i.e., quality of network (QoN),

service (QoS), and experience (QoE) (Falkowski-

Gilski & Uhl, 2020). In the remainder of the paper,

we use the acronym QoS understood as quality in a

wide sense.

Correlation is one of the fundamental tools used

in the analysis and processing of data and signals,

e.g., (Schwarzinger, 2013; Ziółkowski & Kelner,

https://orcid.org/0000-0002-1214-1308

https://orcid.org/0000-0002-3902-0784

2016). Correlation analysis allows you to find

similarities and relationships between two variables,

properties, features, signals, or processes.

Correlation methods are also used in QoS research

and methods. For example, (Yu et al., 2021) proposed

a novel algorithm of approximate service

composition, which is based on QoS correlation. Its

task is to determine the optimal path of service

delivery.

The work (Li et al., 2019) proposes an innovative

approach to service selection that not only considers

QoS correlations of services but also accounts for

QoS correlations of user requirements. The proposed

solution is decentralized, which avoids a single point

of failure. The authors of (Li et al., 2019) presented

experimental results that showed the effectiveness of

the developed solution. A similar approach has been

proposed in (Chervenets et al., 2016). On the other

hand, simplified service selection methods are

described in (Wang et al., 2017; Deng et al., 2016).

The solution shown in (Wang et al., 2017) is based on

the QoS correlation of requirements. The authors of

(Deng et al., 2016) proposed the so-called

correlation-aware service pruning (CASP) method.

Zmysłowski, D. and Kelner, J.

Drive Test-based Correlation Assessment of QoS Parameters for Exemplary Measurements Scenario in Suburban Environment.

DOI: 10.5220/0011575800003318

In Proceedings of the 18th International Conference on Web Information Systems and Technologies (WEBIST 2022), pages 497-504

ISBN: 978-989-758-613-2; ISSN: 2184-3252

497

The CASP approach is based on managing QoS

correlations by accounting for all services that may be

integrated into optimal composite services and prunes

services that are not the optimal candidate services.

The authors of (Ahmad et al., 2020) conducted a

correlation analysis between QoS metrics such as

signal strength, average delay, jitter, and average

packet loss. This analysis was performed for real-time

cellular QoS metrics data, and the authors observed a

significant correlation between several QoS

parameters. Simultaneously, they indicate that some

network emulators, e.g., NetEm (Jurgelionis et al.,

2011; Sliwinski et al., 2010), do not consider the fact

that some QoS parameters are correlated. In (Kim &

Choi, 2010; Bae et al., 2009), the correlations

between QoS/QoE metrics in the video signal were

analyzed.

In this paper, we analyze the correlation between

QoS parameters. The presented results of the analysis

are based on a drive-test (i.e., mobile measurements)

carried out in a suburban environment for two mobile

network operators (MNOs). Tests for video live-

streaming from YouTube (YouTube, 2021) were

selected for the analysis. This service was carried out

using the 3G and 4G standards, i.e., Universal Mobile

Telecommunications System (UMTS) and Long

Term Evolution (LTE), respectively.

The remainder of the paper is organized as

follows. Section 2 presents the test-bed, measurement

scenario, and analyzed QoS metrics. Section 3

contains the results of the performed correlation

analysis of QoS parameters. In the final part of the

paper, we show a summary.

2 DRIVE TEST

The correlation assessment of the QoS metrics was

carried out based on measurements in 3G (UMTS)

and 4G (LTE) mobile networks in Poland. The drive

tests were made by Systemics-PAB Group company

on behalf of the Office of Electronic Communications

(UKE), the Polish telecommunications market

regulator (Kruszewski & Malinowski, 2022). The

measurements were carried out in suburban areas,

near Warsaw in the fourth quarter of 2021. Our

analysis is based on data for the selected scenario, i.e.,

video live-streaming from the YouTube service.

Benchmarking video streaming services was

performed with a dedicated mobile application. The

following sections describe the test-bed, scenario, and

measured QoS parameters.

2.1 Measurement Test-bed

The QoS measurements were made by the Systemics-

PAB Group company using the professional Rohde &

Schwarz test-bed. The measurement system included

SwissQual Diversity Smart Benchmarker Rel. 20.3

based on measuring terminals with SwissQual

QualiPoc software (Kruszewski & Malinowski, 2022).

In tests, the Samsung Galaxy S20 + 5G

(SM-G986BDS) measuring terminals were utilized.

These terminals support the carrier aggregation

technology and all bands used by MNOs in Poland.

A passive scanner, Rohde & Schwarz TSME6,

was used for measuring the QoS metrics that

represented the quality and power of radio signals.

The scanner supports all frequency bands used in

mobile networks.

Figure 1 shows part of the test-bed mounted in the

rear trunk of the car. The measuring terminals were

placed in special housings in the roof box at a height

of about 1.8 m.

Figure 1: Test-bed mounted in the rear trunk (R&S, 2016).

2.2 Measurement Scenario

The research was conducted in the vicinity of

Piaseczno and Góra Kalwaria towns, south of

Warsaw, from November 29 to December 14, 2021.

During the measurements, tests for voice calls and

data transfers were performed. Data transfer tests

included the YouTube live-stream, Hypertext

Transfer Protocol (HTTP) browsing, and transfer. In

this paper, we only analyzed data for the YouTube

test for which three scenarios were performed:

 The G1 scenario played two-minute YouTube

videos with 10-second intervals between

videos.

 The G2 scenario consisted of playing a

60-second movie three times, followed by a

1-second preload ping. Then, there was a

resection pause for 10 seconds.

 The G3 scenario consisted of playing a

10-second movie ten times, followed by a

2-second preload ping. Then, there was a

resection pause for 10 seconds.

QQSS 2022 - Special Session on Quality of Service and Quality of Experience in Systems and Services

498

Figure 2: Maps with measurement points (bins) in YouTube test for a) MNO no. #1 and b) MNO no. #2 (Kruszewski &

Malinowski, 2022).

It is worth emphasizing that the concept of a

measurement point (bin) used in this paper should

refer to a specific section of the measurement route

(as well as the time interval) in which a single test was

carried out for the analyzed scenario.

The tests were conducted for two MNOs

providing telecommunications services in Poland,

marked as MNO no. #1 and #2 – for formal reasons,

the data has been anonymized. In the case of MNO

no. #1, the video transmission service from YouTube

was provided using two technologies, 3G (UMTS)

and 4G (LTE). In the case of MNO no. #2, all

transmissions were made using LTE.

Figure 2 illustrates a situational map with plotted

measurement points (bins) for three scenarios (G1,

G2, and G3) and two MNOs (Kruszewski &

Malinowski, 2022). A green bin means that the test

has been qualified as correct, while other colors

indicate different errors that occurred during the test.

In our analyzes, only the results corresponding to the

green bins (‘Qualified’) were used, i.e., where the

QoS metric values were determined.

2.3 Measured QoS Metrics

During the measurements, for each measurement

point (bin), various parameters were determined

regarding the MNO, time and place of measurement,

used technology (i.e., UMTS or LTE), the average

vehicle speed (AvgSpeed), and QoS parameters,

including:

 VMOS – video mean opinion score – a

subjective metric representing video quality;

the parameter value is determined in the range

from 1 to 5, with an accuracy of 0.1;

 AR – average resolution of video related to the

number of pixels (p) in the film frame level –

parameter representing video quality; its value

is determined in the range from 360p to 1080p

(i.e., Full High Definition);

 RSCP – reference signal code power –

parameter representing radio signal power in

dBm, designated only for UMTS technology;

 EC/IO – downlink carrier-to-interference ratio

– parameter representing radio signal quality in

dB determined for UMTS technology;

 RSRP – reference signal received power –

parameter representing radio signal power in

dBm, designated only for LTE technology;

 SINR - signal-to-interference-plus-noise ratio –

parameter representing radio signal quality in

dB determined for LTE technology.

A description of the last two mentioned metrics is

presented in (Afroz et al., 2015). In this case, the

authors show also their measurement results in the

LTE network.

Drive Test-based Correlation Assessment of QoS Parameters for Exemplary Measurements Scenario in Suburban Environment

499

3 CORRELATION ASSESSMENT

3.1 Correlation Metrics

In our analyzes, we use the Pearson correlation

coefficient (PCC) defined as















PCC X Y











(1)

where





x

and





Yy

represent sets of two

analyzed parameters,





E 

is the expectation

operator,





, and



represent mean

values and standard deviations of

and

respectively, i.e.,



































(2)

PCC is used to assess the correlation between two

QoS metrics or the correlation of the average vehicle

speed (AvgSpeed) with the selected QoS parameter.

In the case of parameters for which PCC > 0.5, the

linear regression using the least squares method and

the deviation of the measurement results in relation to

the obtained line were determined.

3.2 Preparation of Measurement Data

In the YouTube tests performed by Systemics-PAB

Group, total bins (i.e., measurement points) were

equal to 5566 and 6060 bins for MNOs no. #1 and #2,

respectively. In the case of MNO no. #1, video

transmissions using the UMTS and LTE technologies

were performed for 339 (6.1%) and 5227 bins

(93.9%), respectively. For further analysis, only the

data with the ‘Qualified’ status were used, which

means that the test was successful, and the values of

all analyzed metrics were determined. In the case of

MNO no. #1, video transmission tests were successful

with UMTS and LTE for 60.2% and 89.4%,

respectively. In the case of MNO no. #2, video

transmission tests were finished successfully in LTE

technology for 93.3%.

Correlation analyzes were carried out for three

data sets, i.e., MNO#1-UMTS, MNO#1-LTE, and

MNO#2-LTE.

3.3 Correlation between QoS

Parameters

Using the prepared data from tests in UMTS

technology, the PCCs between the four QoS metrics

described in Section 2.3 were determined. For the

MNO#1-UMTS dataset, RSCP, EC/IO, AR, and

VMOS were used. The determined PCC values are

included in Table 1.

Table 1: PCCs between QoS metrics for MNO#1-UMTS.

Metric RSCP EC/IO AR VMOS

RSCP 1.000 0.580 0.242 0.148

EC/IO 1.000 0.395 0.243

AR 1.000 0.546

VMOS 1.000

The obtained results show a significant

correlation between RSCP and EC/IO (PCC = 0.58)

and also between VMOS and AR (PCC = 0.546). For

these parameter pairs, the measurement results are

illustrated in the graphs in Figures 3 and 4,

respectively. In addition, linear regressions have been

determined, which are marked in red.

Figure 3: RSCP versus EC/IO for MNO#1-UMTS.

The following equation describes regression line

between RSCP and EC/IO:



 

dB 1.267 dB 84.5 dBRSCP EC IO 

)

The spread of the empirical RSCP values in

relation to the value determined by the line is defined

by the standard deviation, which is

5.4 dB

RSCP



 .

QQSS 2022 - Special Session on Quality of Service and Quality of Experience in Systems and Services

500

Figure 4: VMOS versus AR for MNO#1-UMTS.

In the case of the relationship between VMOS and

AR for MNO#1-UMTS, the regression line

is defined as



0.0017 p 2.5VMOS AR

(4)

In this case, the deviation of the VMOS results is

equal to

0.5

VMOS



 .

Table 2 shows the results of the correlation

analysis for MNO#1-LTE. In the case of LTE

technology, RSRP and SINR were determined

instead of the RSCP and EC/IO metrics, respectively.

Table 2: PCCs between QoS metrics for MNO#1-LTE.

Metric RSRP SINR AR VMOS

RSRP 1.000 0.647 0.369 0.269

SINR 1.000 0.405 0.298

AR 1.000 0.637

VMOS 1.000

The obtained results show that for MNO#1-LTE,

analogously to MNO#1-UMTS, signal parameters

(i.e., quality – SINR and power – RSRP) and quality

video (i.e., AR and VMOS) are correlated with each

other (i.e., PCC > 0.5). These parameters are

illustrated in Figures 5 and 6, respectively.

The regression line between RSRP and SINR for

MNO#1-LTE is given by the equation:

 

dB 0.776 dB 106.5 dBRSRP SINR 

(5)

In this case, the standard deviation of RSRP

relative to the regression line is

7.0 dB

RSRP



 .

Figure 5: RSRP versus SINR for MNO#1-LTE.

Figure 6: VMOS versus AR for MNO#1-LTE.

The regression line for AR and VMOS metrics

and MNO#1-LTE is as follows:



0.0017 p 2.5VMOS AR

(6)

and the deviation of VMOS is equal to 0.4

VMOS





An analogous analysis of measurement data was

performed for the MNO#2-LTE set. PCCs are shown

in Table 3 and in Figures 7 and 8 for RSRP(SINR)

and VMOS(AR), respectively.

The obtained results indicate the correlation

between SINR and RSRP parameters and a slightly

lower correlation between AR and VMOS (i.e., PCC

= 0.446) than in the case of MNO#1-LTE.

Drive Test-based Correlation Assessment of QoS Parameters for Exemplary Measurements Scenario in Suburban Environment

501

Table 3: PCCs between QoS metrics for MNO#2-LTE.

Metric RSRP SINR AR VMOS

RSRP 1.000 0.657 0.228 0.157

SINR 1.000 0.246 0.156

AR 1.000 0.446

VMOS 1.000

Figure 7: RSRP versus SINR for MNO#2-LTE.

Figure 8: VMOS versus AR for MNO#2-LTE.

The regression line between RSRP and SINR for

MNO#2-LTE is as follows:

 

dB 0.728 dB 105.0 dBRSRP SINR 

(7)

and the RSRP deviation is 6.5 dB

RSRP



 .

For MNO#2-LTE, the regression line between

VMOS and AR is described by the formula:



0.0012 p 3.0VMOS AR

(8)

In this case, the standard deviation of VMOS is

0.3

VMOS





Based on the obtained results, we can see that the

signal parameters represent power (i.e., RSCP and

RSRP for UMTS and LTE, respectively) and quality

(i.e., EC/IO and SINR for UMTS and LTE,

respectively) are strongly correlated with each other.

For the three analyzed data sets, there are PCC > 0.5.

Therefore, regression lines could be derived for these

QoS metrics. It is worth noting that the deviation of

the results of the parameter representing the signal

strength (i.e., RSCP or RSRP) is about 10% of its

variability range, e.g., for MNO#1-LTE,

7.0 dB

RSRP





, while the RSRP varied from

128.6 dBm



58.6 dBm



, which gives a change

range of

70 dB

. Thus, the regression line describes

the relationship between these QoS metrics relatively

well. The obtained equations can be used to determine

the power parameter based on the signal quality

parameter using the normal distribution with an

appropriate standard deviation.

For the video quality parameters, PCC > 0.5 is for

the two sets, i.e., MNO#1-UMTS and MNO#1-LTE,

while for MNO#2-LTE, PCC was slightly below 0.5.

In these cases, the regression lines were also

determined. For MNO#1-UMST and MNO#1-LTE,

line equations are described with identical

coefficients. The difference is for the VMOS standard

deviation values – for UMTS, it is slightly greater

than for LTE. In the case of MNO#2-LTE, the

regression line coefficients are slightly different to

MNO#1-LTE, while VMOS deviation is the smallest.

The obtained regression lines make it possible to

determine the approximate or model VMOS value

based on the average resolution adjusted to the

current link throughput in the radio channel.

Moreover, the results obtained in the tables show

a specific correlation between AR and the radio signal

parameters at the level of 0.22 < PCC < 0.41.

Therefore, the regression lines for such cases may

allow the estimation of AR values based on the signal

parameter (e.g., SINR or RSRP) measurements or

vice versa, which may find practical application.

3.4 Impact of Velocity on QoS

Parameters

Figure 9 shows the velocity distribution for the three

analyzed data sets. We can see that measurements in

suburban areas were usually carried out at speeds

below 40 km/h, while driving at speeds above 50

km/h took place outside the built-up area.

QQSS 2022 - Special Session on Quality of Service and Quality of Experience in Systems and Services

502

Figure 9: Average speed distribution for (a) MNO#1-

UMTS, (b) MNO#1-LTE, and (c) MNO#2-LTE.

In the conducted analysis, we decided to check

whether there is a correlation between the average

vehicle speed and the QoS metrics. The determined

PCCs are included in Table 4.

Table 4: PCCs between QoS metrics and average speed.

Metrics

MNO#1-UMTS MNO#1-LTE MNO#2-LTE

RSCP 0.014

EC/IO 0.031

RSRP 0.069

–

0.004

SINR

–

0.001 0.018

AR 0.046

–

0.021

–

0.039

VMOS 0.068

–

0.026

–

0.030

The obtained results indicate a negligible influence of

speed on the analyzed QoS metrics. The practice of

using data transmission in mobile networks while

traveling by vehicles (e.g., car, train, etc.) shows that

at high speeds, the use of this service type is

problematic. On the other hand, the obtained results

do not indicate a significant impact of speed on the

QoS. Thus, we conclude that the vehicle speed may

have a more substantial effect on session breakup

during video transmission or reconnection between

the mobile terminal and base station. In this paper, we

do not analyze the dataset that was qualified as

‘Failed’ (only ‘Qualified’ data was analyzed).

4 CONCLUSIONS

In this paper, we have presented a correlation analysis

of QoS metrics determined in the measurement

campaign near Warsaw for two MNOs. The study

considered a scenario with the YouTube live-stream

service. The determined PCCs for the parameter pairs

indicated a strong relationship between the signal

parameters – radio signal power and quality, and

between the video quality metrics, i.e., AR and

VMOS. Regression lines for these pairs and metrics

deviations were determined. We also assessed the

impact of car speed on the QoS metrics. In this case,

the obtained PCC values do not show a strong

correlation between them.

It is worth noting that the percentage of

measurements considered for analysis for LTE was

about 90%, while for UMTS is only about 60% (see

Section 3.2). In the UMTS case, the rejection of a

significant amount of data may result from the fact

that older generation solutions (UMTS) cope worse

than LTE technology with user mobility (i.e., Doppler

effect mitigation), ensuring continuity of broadband

transmission, especially when user handover between

the next base stations. Hence, the trend, noticeable in

numerous MNOs, consisting of the 3G technology

abandonment and the use of its spectral resources in

favor of efficient 4G and 5G seems to be justified. On

the other hand, the smaller number of measurement

data used for the UMTS correlation analysis should

insignificantly affect the reliability of the obtained

results, which are close to the LTE results.

In the near future, the authors are planning to

analyze other scenarios carried out in the

measurement campaign and assess the possible

influence of vehicle speed on session interruption.

The obtained results of the analysis will be used in the

method of assessing mobile networks in terms of

QoS/QoE performance assessment developed by the

authors (Zmysłowski & Kelner, 2022).

ACKNOWLEDGEMENTS

This work was financed by the Military University of

Technology under Research Project no. UGB/22-

740/2022/WAT on “Modern technologies of wireless

communication and emitter localization in various

system applications”.

The authors would like to thank the President of

UKE, Dr. Jacek Oko, for providing the measurement

data made by the Systemics-PAB Group company,

which implemented QoS assessment campaigns in

mobile networks on behalf of UKE.

Drive Test-based Correlation Assessment of QoS Parameters for Exemplary Measurements Scenario in Suburban Environment

503

REFERENCES

Afroz, F., Subramanian, R., Heidary, R., Sandrasegaran, K.,

& Ahmed, S. (2015). SINR, RSRP, RSSI and RSRQ

measurements in long term evolution networks.

International Journal of Wireless & Mobile Networks,

7(4), 113–123.

Ahmad, N., Wahab, A., & Schormans, J. (2020).

Importance of cross-correlation of QoS metrics in

network emulators to evaluate QoE of video streaming

applications. In 2020 11th International Conference on

Network of the Future (NoF), 43–47. https://doi.org/

10.1109/NoF50125.2020.9249129

Bae, J., Park, H.-S., Chin, Y. M., & Park, H.-C. (2009). An

identification of correlation between network QoS and

video quality index for IPTV service. In 2009 11th

International Conference on Advanced Communication

Technology (ICACT), 02, 1399–1402.

Chervenets, V., Romanchuk, V., Beshley, H., & Khudyy,

A. (2016). QoS/QoE correlation modified model for

QoE evaluation on video service. In 2016 13th

International Conference on Modern Problems of

Radio Engineering, Telecommunications and

Computer Science (TCSET), 664–666. https://doi.org/

10.1109/TCSET.2016.7452146

Deng, S., Wu, H., Hu, D., & Leon Zhao, J. (2016). Service

selection for composition with QoS correlations. IEEE

Transactions on Services Computing, 9(2), 291–303.

https://doi.org/10.1109/TSC.2014.2361138

Falkowski-Gilski, P., & Uhl, T. (2020). Study of cellular

network quality on the eve of 5G technology on a

selected example [in Polish: Badanie jakości sieci

komórkowych w przededniu technologii 5G na

wybranym przykładzie]. Telecommunication Review

and Telecommunication News, 93(7–8), 338–341.

https://doi.org/10.15199/59.2020.7-8.45

Jurgelionis, A., Laulajainen, J.-P., Hirvonen, M., & Wang,

A. I. (2011). An empirical study of NetEm network

emulation functionalities. In 2011 20th International

Conference on Computer Communications and

Networks (ICCCN), 1–6. https://doi.org/10.1109/

ICCCN.2011.6005933

Kim, H. J., & Choi, S. G. (2010). A study on a QoS/QoE

correlation model for QoE evaluation on IPTV service.

In 2010 12th International Conference on Advanced

Communication Technology (ICACT), 2, 1377–1382.

Kruszewski, R., & Malinowski, J. (2022). UKE Poland

measurement specification (1.7). Systemics-PAB Sp. z

o.o. for Office of Electronic Communications.

Li, D., Ye, D., Gao, N., & Wang, S. (2019). Service

selection with QoS correlations in distributed service-

based systems. IEEE Access, 7, 88718–88732.

https://doi.org/10.1109/ACCESS.2019.2926127

Lloyd-Evans, R. (2002). QoS in integrated 3G networks.

Artech House Publishers.

Oodan, A., Ward, K., Savolaine, C., Daneshmand, M., &

Hoath, P. (2002). Telecommunications quality of

service management: From legacy to emerging

services. The Institution of Engineering and

Technology.

Razeghi, M. (2007). End-to-end quality of service over

cellular networks: Data services performance

optimization in 2G/3G (G. Gomez & R. Sanchez, Eds.).

John Wiley & Sons.

R&S. (2016). Benchmarker II takes off into the Polish

telecommunications market (PD 5214.6774.32). Rohde

& Schwarz.

Schwarzinger, A. (2013). Digital signal processing in

modern communication systems. Andreas Schwarzinger.

Sliwinski, J., Beben, A., & Krawiec, P. (2010). EmPath:

Tool to emulate packet transfer characteristics in IP

network. In F. Ricciato, M. Mellia, & E. Biersack

(Eds.), In 2010 International Workshop on Traffic

Monitoring and Analysis (TMA) (pp. 46–58). Springer.

https://doi.org/10.1007/978-3-642-12365-8_4

Wang, Y., He, Q., Ye, D., & Yang, Y. (2017). Service

selection based on correlated QoS requirements. In

2017 IEEE International Conference on Services

Computing (SCC), 241–248. https://doi.org/10.1109/

SCC.2017.38

YouTube. (2021). https://www.youtube.com/

Yu, J., Lin, Z., Yu, Q., & Xiao, X. (2021). QoS correlation-

based service composition algorithm for multi-

constraint optimal path. In 2021 11th International

Conference on Information Technology in Medicine

and Education (ITME), 117–121. https://doi.org/

10.1109/ITME53901.2021.00034

Ziółkowski, C., & Kelner, J. M. (2016). Influence of

receiver/transmitter motion direction on the

correlational and spectral signal properties. In 2016

10th European Conference on Antennas and

Propagation (EuCAP), 1–4. https://doi.org/10.1109/

EuCAP.2016.7481225

Zmysłowski, D., & Kelner J. M. (2022) The concept of

application of the Wroclaw taxonomy for QoS

assessment in mobile networks. In 2022 18th

International Conference on Web Information Systems

and Technologies (WEBIST), Special Session on

Quality of Service and Quality of Experience in Systems

and Services (QQSS).

QQSS 2022 - Special Session on Quality of Service and Quality of Experience in Systems and Services

504