11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

Feasibility Study of Universitas Advent Indonesia
Network for Hybrid Distance Learning
Jay Idoan Sihotang1*
Universitas Advent Indonesia
jay.sihotang@unai.edu

ABSTRACT
The rapid development of the internet has spurred many fields to adopt remote meetings.
Indonesia Adventist University (UNAI) has the capability to implement hybrid learning,
particularly in distance learning. In this research, testing will be conducted at several student
testbed locations across UNAI, focusing on various QoS (Quality of Service) parameters,
including bandwidth, delay, jitter, and UDP packet loss rate. The results show significant
variations in QoS across different testbeds, presenting specific challenges for future
experimental plans.
Keywords: Network Readiness, QOS, Hybrid Learning, Distance Learning

INTRODUCTION
The rapid development of technology has allowed many fields to operate more freely
and reach an increasing number of users via the Internet. One of the triggers for the rapid
business transition in adopting technology was the 2020 pandemic. This swift transition also
occurred at Universitas Advent Indonesia, where a strategic decision had to be made regarding
the adoption of technology for distance learning (Limbong, 2022). Some approaches taken
included using Zoom for meetings and Moodle as the Learning Management System. In several
previous studies, it has been proven that Zoom has been instrumental in facilitating teaching
and learning activities (Boy et al., 2022; Iryani et al., 2020).
Some challenges encountered during distance learning include interrupted or hindered
media transmission due to varying quality of service in different regions. In previous research
conducted by (Sihotang & Bandung, 2016), Background traffic has a significant impact on the
quality of multimedia services, especially for online classes that use the UDP protocol with a
higher bitrate compared to text-based services that use the TCP protocol. This prompted
researchers, at the end of the pandemic, to conduct a study on the readiness of Universitas
Advent Indonesia to handle hybrid distance learning (offline classes with some attendees
joining from another city via online meetings). This research involved several testbeds

1532

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

distributed from the western to eastern parts of Indonesia, and it is expected to provide a
comprehensive overview of the network conditions at those testbed locations.

METHODS
In conducting the research, the following are the research stages used:
Literature
Study

Choosing
Parameters

Creating
Testbed
Testing
Scenario

Acquiring
data from
Various
Testbed

Analyzing the
Collected
data

Summarized
for Next
Phase

Figure 1 Research Methods
The initial stage begins with studying several pieces of literature from previous
research, which are summarized into the background that motivates the study. This is then
distributed across the document in the form of definitions or explanations for each research
instrument.
QOS Parameters And Testbed Scenario
In collecting data from 10 testbeds spread across various locations in Indonesia,
including Bandung, Jakarta, Bekasi, East Java, Kalimantan, Pekanbaru, North Sumatra, and
Kupang. The data was collected remotely using AnyDesk to connect to volunteer computers at
those locations. Several Quality of Service (QoS) parameters were used in this research. These
QoS parameters include Bandwidth, Delay, Jitter, and UDP Packet Loss Rate (PLR) (Sigeti et
al., 2013).
Each QoS test followed the scenarios below:
1. Data was collected on four different weekdays (08:00 AM – 05:00 PM) during periods
when students were attending online classes via Zoom.
2. Additional data was collected during non-peak periods over four time periods on the
same days during the weekend (or other days when students had no classes).
The measurement of QoS was conducted using Iperf/Jperf and PING (Packet Internet Gopher).
Below is the explanation of the scenario for measuring each QoS parameter.
1. Bandwidth
Bandwidth refers to the maximum data transfer rate of a network or internet connection,
indicating how much data can be sent or received per second. It is typically measured
in bits per second (bps), with higher bandwidth allowing more data to flow through the
connection. While higher bandwidth can improve speed and performance, it is just one
factor in overall network quality, alongside latency and reliability. Bandwidth capacity
was measured using TCP traffic. TCP datagrams were sent using the Iperf/Jperf tool.
To run this scenario, one host was needed as a client connected to the DSL internet
1533

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

network and one server located at UNAI. Each measurement period was conducted for
10 seconds. The Iperf server could be accessed starting at 08:00 AM and was turned off
at 05:00 PM. Volunteers at each testbed could perform the tests according to the
scenario after confirming that they were ready to conduct the testing.
2. Delay
Delay refers to the time it takes for data to travel from the source to the destination in a
network. It is typically measured in milliseconds (ms) and can be affected by factors
like distance, network congestion, and the number of intermediate devices (such as
routers) along the path. High delay, often called latency, can cause noticeable lags in
communication, especially in real-time applications like video calls or online gaming.
E2E Delay is measured using ICMP request and reply to packets (Ping) (ITU, 1994).
To obtain more accurate data, the measurement is performed several times with
different packet sizes. The packet sizes sent are 1350 bytes (average Max MTU of the
Testbed network) and 40 bytes (as referenced in the Dischinger paper). Each time the
measurement is performed, 100 packets are sent. Like bandwidth, volunteers in each
testbed can conduct tests according to the scenario after confirming their participation
in the testing.
3. Jitter & UDP Packet Loss Rate
Jitter refers to the variation in packet arrival times in a network, meaning that data
packets may arrive at irregular intervals rather than a steady, predictable rate. And UDP
Packet Loss Rate is the percentage of data packets lost during transmission over a
network using the User Datagram Protocol (UDP) (Bekele et al., 2024). A high value
of Jitter and/or UDP Packet Loss Rate can cause disruptions in real time
communications. Jitter and UDP packet loss rate measurements are conducted using
UDP traffic. UDP datagrams are sent using Iperf tools. To run this scenario, 2 hosts are
required: one acts as the server (located at UNAI) and the other as the client (located at
each testbed). Each measurement is performed for 10 seconds, and each measurement
period lasts for 10 seconds. The Iperf server is accessible from 08:00 to 17:00.
Volunteers in each testbed can perform the tests according to the scenario after
confirming their participation in the testing.
Each set of data collected from the sessions is presented in table form to directly assess
the quality of service for each testbed. The presentation of the measurement results from each
testbed will be discussed in the next section.

RESULTS AND DISCUSSION
The results from the field tests or testbed, conducted at the end of 2021 with the
assistance of several student volunteers who were willing to be remotely managed, and
1534

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

following various specific scenarios and stages for each tested parameter, are presented in the
following section.
Bandwidth
The bandwidth testing was done in various time during each Weekday or Weekend
scenario as mentioned in previous section. Bandwidth is a primary parameter commonly
understood by the public as a measure of network quality. This has led to the paradigm among
the public that the higher the bandwidth, the better the ISP network. Bandwidth testing is
conducted over several periods, with each testing period consisting of 4 sequences. The Table
1 below refer to the Bandwidth Testing of each testbed in Weekdays (the test performed while
student having their online meeting using zoom).
Table 1 Bandwidth – On Class (Mbps)
Bandwidth - On Class (Busy)
Testbed
Seq. 1
1
3,55
2
1,27
3
4,5
4
30,5
5
3,59
6
25,7
7
4,3
8
7,26
9
2,41
10
2,28

Seq. 2
3,1
0,973
3,66
28,8
1,9
26,5
3,73
3,09
2,45
0,921

Seq. 3
3,37
1,26
0,419
33,5
3,71
26
3,92
3,38
0,349
0,587

Sec. 4
3,77
1,21
3,44
33,4
3,47
25,2
2,23
5,3
3,23
0,944

Average
3,45
1,18
3,00
31,55
3,17
25,85
3,55
4,76
2,11
1,18

And now, to also measure the available bandwidth on weekends or during off-class
periods, the same test is conducted to get an overview of the bandwidth availability during free
time or when no Zoom classes are being held. Table 2 below provides an overview of the
available bandwidth during off-class times or weekends.
Table 2 Bandwidth – Off Class (Mbps)
Bandwidth - Off Class (Free)
Testbed
Seq. 1
1
3,67
2
2,83
3
1,47
4
33,5
5
3,74
6
25,3
7
4,29
8
7,65
9
3,1
10
0,807

Seq. 2
3,98
1,62
1,36
33,4
3,34
25,1
4,3
11
3,75
2,31

Seq. 3
3,35
2,25
1,68
33,4
3,28
26,5
4,3
7,74
2,48
2,31

Sec. 4
4,09
0,639
1,47
33,4
2,75
26,4
4,27
8,53
3,39
2,35

Average
3,77
1,83
1,50
33,43
3,28
25,83
4,29
8,73
3,18
1,94

When comparing the bandwidth availability during periods of Zoom meetings or when
no classes are being held, there are several differences. This also does not account for
background traffic that may occur during testing due to other users accessing the network at
1535

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

the same time. The highest and most consistent bandwidth was found in Testbed 4 located in
Bekasi, while the lowest bandwidth was recorded in Testbed 10 in Sidoarjo (followed by
Testbed 2 located in North Sumatra). Figure 2 provides an infographic showing the differences
in bandwidth between the two testing periods.
Bandwith Comparison (Mbps)
33,43
25,83

31,55
25,85
8,73
3,77
3,45

4,29
3,55

3,28
3,17

1,50
3,00

1,83
1,18

3,18
2,11

4,76

1,94
1,18

1 (Jakarta)

2 (Medan)

3 (Bandung)

4 (Jakarta)

5
(Pekanbaru)

6 (Bekasi)

7
(Balikpapan)

Bandwidth - Off Class

3,77

1,83

1,50

33,43

3,28

25,83

4,29

8,73

3,18

1,94

Bandwidth - On Class

3,45

1,18

3,00

31,55

3,17

25,85

3,55

4,76

2,11

1,18

8 (Kupang) 9 (Surabaya) 10 (Sidoarjo)

Figure 2 Bandwidth Comparison (Mbps)
Delay
Delay testing is conducted after the iperf command is executed in the bandwidth test,
following each testing sequence in different periods. The packet size values used are 40 bytes
(default) and 1350 bytes (Average Max MTU). Table 3 below shows the differences in delay
between testing scenarios during Zoom classes (On Class) and when no classes are being held
or during weekends (Off Class) for each TestBed (TB).
Table 3 Delay Comparisons (ms)
Delay (40) On Class
Delay (1350)
- On Class
Delay (40) Off Class
Delay (1350)
- Off Class

TB-1

TB-2

TB-3

TB-4

TB-5

TB-6

TB-7

TB-8

TB-9

TB-10

Average

9,25

118

130

6

49,75

8,75

64,5

151,25

149,75

444,75

113,2

10,5

140,5

163,25

7,25

53,25

9,75

66

152,75

192,75

381

117,7

7

153,75

118

6,75

47,75

12

39

96

115

71,25

66,65

8

158,25

157,75

8

49,5

17

41,75

119,5

126,5

69,75

75,6

From the table above, it can be concluded that, in general, the level of delay increases
when testing is conducted during class hours using Zoom meetings. Although testing outside
class hours is not free from other traffic on the network that directly affects delay values. On
average, the delay values in each testbed still fall within a good classification (below 150 ms)
(International Telecommunication Union, 2001). However, it is noted that some testbeds have
unacceptable delay values, especially when compounded by the load of meetings conducted
via Zoom.
1536

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

Delay Comparison (ms)
500,00

444,75

450,00
381,00

400,00
350,00
300,00
250,00
200,00
150,00

192,75
152,75 149,75
151,25
126,50
119,50 115,00
96,00

163,25
158,25
157,75
153,75
140,50
130,00
118,00
118,00

100,00

66,00
64,50
53,25
49,75
49,50
47,75
41,75
39,00
17,00
12,00
9,75
8,75
8,00
7,25
6,75
6,00

50,00 9,25
10,50
8,00
7,00

117,70
113,20
75,60
71,25
69,75 66,65

0,00

Delay (40) - On Class (Busy)

Delay (1350) - On Class (Busy)

Delay (40) - Off Class (Free)

Delay (1350) - Off Class (Free)

Figure 3 Delay Comparison (ms)
Jitter
Immediately after measuring the delay for both 40 and 1350 bytes, jitter measurements
were conducted using the UDP protocol at 256K. The tests, performed during different
periods—class time and off-class time—showed differences consistent with the previous two
parameters tested. The testing was still conducted over 4 sequences in each testing period.
Table 4 below summarizes the jitter testing for each scenario in each TestBed (TB).
Table 4 Jitter Comparison (ms)
Jitter –
UDP - 256K –
On Class
Jitter –
UDP - 256K Off Class

TB-1

TB-2

TB-3

TB-4

TB-5

TB-6

TB-7

TB-8

TB-9

TB-10

Average

0,57

58,16

28,30

0,07

19,44

2,13

23,13

15,78

26,05

0,46

17,41

0,48

22,09

21,71

0,07

12,18

0,52

11,98

9,13

18,24

0,52

9,69

Consistent with the previous results, the condition of the network when burdened with
online classes via Zoom increases the jitter values. However, on average, the jitter values are
still considered acceptable (International Telecommunication Union, 2001). It is noted that
Testbed 2 has a high jitter value that falls into the unacceptable category.
1537

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

Jitter Comparison (ms)
70,00
60,00
50,00
40,00
30,00
20,00
10,00
0,00

58,16

22,09

28,30
21,71

0,57
0,48

23,13
11,98

19,44
12,18

15,78
9,13

26,05
18,24

2,13
0,52

0,07
0,07

Jitter - UDP - 256K - On Class (Busy)

0,52
0,46

Jitter - UDP - 256K - Off Class (Free)

Figure 4 Jitter Comparison (ms)
UDP Packet Loss Rate
UDP Packet Loss Rate measures the number of datagrams received from each data
transmission session. UDP is the protocol used for transmitting both static and dynamic images.
Below is a comparison of the UDP Packet Loss Rate from each TestBed (TB) for the On Class
and Off Class scenarios. One period of the scenario consists of 4 testing sequences.
Table 5 UDP Packet Loss Rate Comparison (%)
Packet Loss Rate
- UDP - 256K On Class
Packet Loss Rate
- UDP - 256K Off Class

TB-1

TB-2

TB-3

TB-4

TB-5

TB-6

TB-7

TB-8

TB-9

TB-10

Average

10%

49%

1%

32%

8%

0%

27%

33%

20%

57%

24%

16%

14%

1%

17%

20%

0%

35%

0%

25%

55%

18%

From the test results above, the overall UDP packet loss rate is relatively high,
especially in several testbeds with low bandwidth. In contrast, some testbeds with high
bandwidth (e.g., TB-3, TB-6) experience very minimal datagram loss. Meanwhile, Testbed-10
demonstrates poor performance, with a UDP Packet Loss Rate exceeding 50% (Bekele et al.,
2024).

1538

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

UDP Packet Loss Rate Comparison (%)
60%

57%55%

49%

50%
40%
30%
20%

35%
27%

32%
16%
10%

17%

14%

10%

33%
25%
20%

20%
8%

1% 1%

0% 0%

0%

0%

Packet Loss Rate - UDP - 256K - On Class (Busy)

Packet Loss Rate - UDP - 256K - Off Class (Free)

Figure 5 UDP Packet Loss Rate Comparison (%)

CONCLUSION
Driven by the curiosity about the readiness of each student’s network at Advent
University Indonesia and supported by a testing scenario encompassing several testing
parameters, this study provides an overview of students’ network conditions. In general, most
classes conducted via Zoom meetings run smoothly. However, some students have difficulty
receiving good video quality, as reflected in several testbeds with UDP Packet Loss Rates
above 30%. Nevertheless, the audio from conversations during Zoom meetings remains clear,
supported by delay values below 150 ms. Given the varying bandwidths among testbeds, it
indicates that Universitas Advent Indonesia needs to consider several recommendations for the
implementation of hybrid learning: 1) Ensuring that each student attending virtual meetings
has adequate bandwidth (especially in the selection of ISPs (Pranata & Rizki Dewantara,
2024)), 2) Considering the use of WebRTC in optimizing virtual meeting technology (Amaral
et al., 2014), 3) Creating a network prioritization scale on campus to reduce the likelihood of
significant delays due to insufficient bandwidth in classroom environments.
For researchers, the subsequent steps that can be undertaken involve the development
of experimental scenarios designed to simulate real-time communication between testbeds
using external applications. These simulations are critical for exploring various factors that
influence the quality of online learning environments. Specifically, researchers might examine
the effects of different video bitrates, which can significantly impact the clarity and fluidity of
video streams during virtual classes. Additionally, varying levels of background traffic will be
assessed to understand how concurrent data usage affects communication quality. Another
important aspect of these scenarios is optimizing network priorities. This involves creating
strategies that allocate bandwidth more effectively, thereby minimizing disruptions and delays
1539

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

that could hinder the learning experience. Customized bitrates for each testbed will also be
determined, allowing for tailored solutions that cater to the unique characteristics of each
environment.
The aim of these experimental simulations is to provide insights that can guide the
effective integration of technology in hybrid learning models. By analyzing the results,
educators and institutions can better understand how to leverage both on-campus networks and
rural connectivity to enhance educational outcomes. This knowledge is vital in ensuring that
all students have equitable access to high-quality learning experiences, regardless of their
location. Through comprehensive testing and analysis, the research aims to contribute
significantly to the field of educational technology.

1540

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in
Shaping Today's Generation Decision Making and Building
Connections” October 22-23, 2024

REFERENCES
Amaral, V., Lima, S. R., Mota, T., & Chainho, P. (2014). Exploring WebRTC Technology for
Enhanced Real-Time Services. In Á. Rocha, A. M. Correia, F. . B. Tan, & K. . A.
Stroetmann (Eds.), New Perspectives in Information Systems and Technologies,
Volume 2 (Vol. 276, pp. 43–52). Springer International Publishing.
https://doi.org/10.1007/978-3-319-05948-8_5
Bekele, B. E., Tokarz, K., Gebeyehu, N. Y., Pochopień, B., & Mrozek, D. (2024). Performance
Evaluation of UDP-Based Data Transmission with Acknowledgment for Various
Network Topologies in IoT Environments. Electronics, 13(18), 3697.
https://doi.org/10.3390/electronics13183697
Boy, R., Sihotang, J. I., & Moedjahedy, J. (2022). Analysis of the Effectiveness of Online
Learning Using Zoom Meetings Using COBIT 5 Framework with Deliver, Service, and
Support
(DSS).
CogITo
Smart
Journal,
8(1),
35–46.
https://doi.org/10.31154/cogito.v8i1.389.35-46
International Telecommunication Union. (2001). ITU-T G.1010 SERIES G: TRANSMISSION
SYSTEMS AND MEDIA, DIGITAL SYSTEMS AND NETWORKS End-user Multimedia
QoS Categories (ITU-T G.1010; p. 18). International Telecommunication Union.
https://www.itu.int/rec/T-REC-G.1010-200111-I
Iryani, N., Dwi, A., & Masykuroh, K. (2020). Analysis of QoS Performance of Online Learning
Applications during Work Hours. JTERA (Jurnal Teknologi Rekayasa), 5(2), 201.
https://doi.org/10.31544/jtera.v5.i2.2020.201-206
ITU, T. S. S. of. (1994). ITU-T Recommendation G.114: Transmission Systems and Media:
General Recommendations on the Transmission Quality for an Entire International
Telephone
Connection:
One-Way
Transmission
Time.
International
Telecommunication Union. https://books.google.co.id/books?id=eQ9RHwAACAAJ
Limbong, A. (2022). Students’ Perception onto Online Learning Management System in
Universitas
Advent
Indonesia.
TeIKa,
12(01),
1–10.
https://doi.org/10.36342/teika.v12i01.2803
Pranata, E. J. & Rizki Dewantara. (2024). Measurement And Analysis Of Quality Of
Service(Qos)4g Network (Telkomsel, Xl, And Indosat Operators). Cyber Security Dan
Forensik Digital, 6(2), 69–75. https://doi.org/10.14421/csecurity.2023.6.2.4246
Sigeti, P., Oravec, M., & Pavlovičová, J. (2013). Some aspects of QoS in wireless networks.
Proceedings ELMAR-2013, 173–176.
Sihotang, J. I., & Bandung, Y. (2016). The impact of background traffic towards multimedia
content delivery Quality on limited capacity network. 2016 10th International
Conference on Telecommunication Systems Services and Applications (TSSA), 1–6.
https://doi.org/10.1109/TSSA.2016.7871082

1541