Capacity Improvement and Protection of LTE Network on Ethernet Based Technique

TELKOMNIKA, Vol.16, No.1, February 2018, pp. 200~209
ISSN: 1693-6930, accredited A by DIKTI, Decree No: 58/DIKTI/Kep/2013
DOI:10.12928/TELKOMNIKA.v16i1.7674  200
Received October 26, 2017; Revised December 1, 2017; Accepted January 10, 2018
Capacity Improvement and Protection of LTE Network
on Ethernet Based Technique
Fadli Sirait*, Akhmad Wahyu Dani, Triyanto Pangaribowo
Department of Electrical Engineering, Universitas Mercu Buana
Jakarta, Indonesia
*Corresponding author, e-mail: fadli.sirait@mercubuana.ac.id
Abstract
High demands for data rates in mobile communications is the reason for developing broadband
wireless access technologies. Long Term Evolution (4G LTE) networks which offer significantly higher data
rates and require suitably higher capacity backhaul networks. To prepare for the high data rates usage in
4G LTE, operators are using ethernet services in terms of backhaul connectivity. Protection packet
switching developed to anticipated network failure on ethernet based network technology. The failures in
the network include the link fails to connect to each network element, the network element fails to transfer
the data to the destination, or the quality drops below the standard. In this paper we used two ethernet
based technique, namely Ethernet over SDH and MPLS-TP with ring protection to anticipated network
failure on these techniques. Furthermore, we measured performance of network by measuring and
comparing the throughput, latency and jitter between Ethernet over SDH and MPLS-TP. We used
bandwidth capacity 240 Mbps as plant bandwidth link and worked in MIMO 2 2. The results of
measurements indicated that MPLS-TP with ring protection is the best technique to enhanced the
performance of LTE network.
Keywords: LTE, MPLS-TP, Ethernet over SDH, Protection Switching, Ring Protection
Copyright © 2018 Universitas Ahmad Dahlan. All rights reserved.
1. Introduction
The 4
th
Generation Long Term Evolution (4G LTE) is an emerging technology to
address the needs of high demand for the data rate. LTE aims to offer a minimum of 100/50
Mbps DL/UL (Downlink/Uplink, 1 sector, 20MHz spectrum) and up to 1 Gbps DL (3 sectors,
DL/UL of 300/150 Mbps per sector) per LTE Base Station (eNB) [1]. Moreover, there is a need
in terms of very high link utilization (<<100 percent) to support the improvement of customer
services such as multimedia distribution, e-Commerce, and cloud computing or gaming systems
as a part of future converged Industrial Internet of Things (IIoT) [2]. Ethernet based technique is
developing technology to support the backhaul capacity of LTE. Despite having such
advantage, there is one big issue in ethernet networks in terms of survivability focusing on
protection switching.
Aleksandra et al [3] have describe that Carrier Ethernet, one of the candidate
technologies for mobile backhaul, protects the network from users that want to ﬂood the network
with their data and manages to keep the delay experienced by other users low. Andrei et al [4]
have analyzed the impact of backhaul packet delay (latency) on the LTE S1-U interface, which
provides user plane transport between the Core Network and the Evolved NodeBs.
Andrei et al [5] have analyzed the impact of backhaul packet loss by focusing on the LTE S1-U
interface, which provides user plane transport between the Core Network (CN) and the Evolved
NodeBs (eNBs). Yi Shi et al [6] introduced a novel wireless backhaul framework for small cell
networks by introducing a new backhauling component, which is referred to as Type-A relay in
order to capable of communicating simultaneously with two macrocells by leveraging the
downlink resource of one cell as well as the uplink resource of the other, which effectively
operates in full-duplex mode without incurring additional costs to acquire carrier resources.
Vijayalakshmy et al [7] presented the analyzing of convergence of 3G (3rd Generation), IEEE
802.16, IEEE 802.11 with an LTE backbone network for improving the network performance.
Christian et al [8] reviews the use case for MPLS-TP ring protection switching, how one of the
proposed mechanisms, namely MPLS-TP Ring Protection Switching (MRPS), can address the

TELKOMNIKA ISSN: 1693-6930 
Capacity Improvement and Protection of LTE Network on Ethernet Based … (Fadli Sirait)
201
requirements to support the packet transport reliability and some possible enhancements to the
MRPS base mechanism. Wenjun et al [9] have proposed a method of ring protection to reduce
the amount of packet loss. Jeong-dong et al [10] have discussed a linear protection switching
mechanism on a mesh-based MPLS-TP network and it is compatible to other types of network
topologies. Chang-Gyu et al [11] have presented a finding that the rapidly developing computer
networking technology, the Software Defined Networking (SDN), can be used to manage MPLS-
TP by using Transport-SDN (T-SDN) controller. Shicheng et al [12] proposed a protection
mechanism in Passive Optical Network (PON) to achieve high availability. Amra at al [13]
presented the result of practical performance benchmark analysis of the state-of-the-art
procedures for restoration and protection of Ethernet over SDH networks. Patteti et al [14]
describe that in LTE there is 100.8 Mbps for single chain, therefore in this paper, we used LTE
system with 2 2 MIMO (2T2R), then the throughput will be two times of single chain throughput.
i.e. 2 x 01.6 Mbps. Patteti et al [13] also describe that there is 25% of overhead used for
controlling and signaling. So the effective throughput will be 151.2 Mbps.
In this paper, we used Ethernet over SDH and MPLS-TP to describe performance of the
LTE network. On the other hand, there is still problem exist in these technique in terms of
network failure, therefore, to overcome this problem, we are implementing ring protection
method on Ethernet over SDH and MPLS-TP based network and then analyzing and comparing
performance of these network. There are four Quality of Service (QoS) parameters that are
tested in this paper, namely throughput utilization, jitter, latency, and recovery time. We used
these parameters to determine that the quality of these network with ring protection is not lower
than the quality of the main network.
The rest of this paper is organized as follows: Ethernet Based Networks is presented in
Section 2, in Section 3 Ring Protection on Ethernet Based Network is describe, in Section 4
Impairment Testing is described, in Section 4 the Results and Analysis are discussed, and in
Section V the Conclusion of this paper is presented.
2. Ring Protection on Ethernet Based Networks
2.1. Ring Protection on Ethernet over SDH
In Figure 1, it can be seen that the ring protection is presented as a red ring. The ring
protection is working when the main ring fails to work. The scenario in this section is the main
link between MUX-SDH-1 and MUX-SDH-3 is disabled manually (indicated with red x marks),
then the packets will travel on the ring protection network automatically. The BER Test Ethernet
is used to measure the QoS on the ring protection network. Then the main ring is enabled again
and the packets switched back to the main ring. The recovery time is the time needed for the
packets to revert back to the main ring. In this step throughput, jitter latency and recovery time
is measured.
Figure 1. Ethernet over SDH with Ring Protection Topology
X

 ISSN: 1693-6930
TELKOMNIKA Vol. 16, No. 1, February 2018: 200 – 209
202
2.2. Ring Protection on MPLS-TP
In Figure 2 it can be seen that the ring protection is presented as a red ring. The ring
protection is working when the main ring fails to work. The scenario in this section is the main
link between NE_MPLS-TP_1 and NE_MPLS-TP_2 is disabled manually (indicated with red x
marks), then the packets will travel on the ring protection network automatically. The BER Test
Ethernet is used to measure the QoS on the ring protection network. Then the main ring is
enabled again and the packets switched back to the main ring. The recovery time is the time
needed for the packets to revert back to the main ring. In this step throughput, jitter latency and
recovery time is measured.
Figure 2. MPLS-TP with Ring Protection Topology
3. Impairment Testing for LTE Networks
Impairments is affecting to traffic packets traversing the LTE networks. The cause for
these impairments are cumulative random events (e.g. busy hours, high profile events, panic
usage, localized service disruptions, dynamic routes and paths for packets, faulty or overloaded
network elements) [14]. RFC 2544 Standard testing are using to validate the Quality of Services
(QoS) in LTE networks. The RFC 2544 is a standard developed by the Internet international
standardization body, the Internet Engineering Task Force (IETF), which contains the necessary
testing methods to measure the quality of an Ethernet network. The parameters that are testing
in [16] are as shown in Tables 1, 2 and 3.
Table 1. Standard for Throughput
Category Throughput
Very Good 76% - 100 %
Good 51% - 75 %
Modearte
Bad
25% - 50 %
< 25%
Table 2. Standard for Jitter
Category Throughput
Very Good 0 ms
Good 0 – 75 ms
Modearte
Bad
76 – 125 ms
126 – 225 ms
Table 3. Standard for Latency
Category Throughput
Very Good t end to end < 150 ms
Good 150 ≤ t end to end ≤
300 ms
Modearte
Bad
t end to end > 450 ms
According to the ITU-T standard, the G 803.2, on Ethernet ring protection switching, the
standard recovery time is less than 50ms.
4. Results and Analysis
In this section, we provide the measurements results to verify the performance of
network quality for all the frame sizes measured. There are seven difference sizes of frame
length that we used during measurement process, which are 64, 128, 256, 512, 1024, 1280,
and 1518 fit to testing standard RFC 2544 [16].

203
4.1. Throughput on the Main Ring Network
Figure 3 presents the test results of throughput parameter values with different frame
length on the main ring, which could be seen that the actual link plan for throughput is 240 Mbps
for all frame sizes. The test results for Ethernet over SDH technique is varied from 224.94 Mbps
to 225.2 Mbps for all frame sizes. Whereas, the test results for MPLS-TP technique is 240 Mbps
for all frame sizes.
Figure 3. The Value of Throughput on the Main Ring for Actual Link Plan, Ethernet over SDH
and MPLS-TP
Figure 4 indicated the throughput utilization of those techniques on the main ring, where
the throughput utilization for Ethernet over SDH varied from 93.72 % to 93.83 % for all frame
sizes, whereas the throughput utilization for MPLS-TP is 100 % for all frame sizes.
Figure 4. The Value of Throughput on the Main Ring in Terms of Utilization in Percentage from
Actual Link Plan, Ethernet over SDH and MPLS-TP
From Figure 3 and Figure 4, we can see that there is gap between Ethernet over SDH
technique and actual link plan in term of throughput values, where the value of throughput from
the Ethernet over SDH technique didn’t reach the defined value that is 240 Mbps. Whereas
there is no gap between MPLS-TP technique and actual link plan.
4.2. Frame Loss on the Main Ring Network
Figure 5 shows the frame loss of Ethernet over SDH on the main ring and MPLS-TP on
the main ring with different frame length. The results showed the frame loss values of Ethernet
over SDH on the main ring varied from 14.8 Mbps to 15.08 Mbps for all frame sizes, whereas,
there is no frame loss in MPLS-TP on the main ring.

 ISSN: 1693-6930
204
Figure 5. The Value of Frame Loss on the Main Ring Network from Actual Link Plan, Ethernet
over SDH and MPLS-TP
In Figure 6 shows the values of frame loss in percentage, where the percentage values
of frame loss in Ethernet over SDH on the main ring varied from 6.17 % to 6.28%, whereas,
there is no frame loss in MPLS-TP on the main ring. All the results showed that on the main ring
performance of MPLS-TP is better than Ethernet over SDH.
Figure 6. The Value of Frame Loss in Percentageon on the Main Ring Network from Actual Link
Plan, Ethernet over SDH and MPLS-TP
4.3. Latency on the Main Ring Network
Figure 7 illustrates the results of measurements for latency values on the main ring of
Ethernet over SDH and MPLS-TP with different frame length. Both of the results is kept below
the standard QoS that is 150ms [17]. However, the results verify that MPLS-TP has better
performance as compared to Ethernet over SDH.
Figure 7. The Latency Values of Ethernet over SDH and MPLS-TP on the Main Ring Network

205
4.4. Jitter on the Main Ring Network
Table 4 shows the result of measurements on the main ring network for both Ethernet
over SDH and MPLS-TP. The results showed that the jitter’s values for both Ethernet over SDH
and MPLS-TP is matched with defined category that us 0-75ms [17]. The values of latency
Therefore, the value of jitter for both Ethernet over SDH and MPLS-TP are available to support
LTE network.
Table 4. Jitter on the Main Ring Network
Maximum
(ms)
Current
(ms)
Average
(ms)
Estimate
(ms)
Category
0.015 0.015 0.015 0.015 0-75 ms
4.5. Throughput on the Ring Protection Network
Figure 8 presents the test results of throughput parameter values with different frame
length, which could be seen that the actual link plan for throughput is 240 Mbps for all frame
sizes. The test results for Ethernet over SDH technique is varied from 224.94 Mbps to 225.2
Mbps for all frame sizes. Whereas, the test results for MPLS-TP technique is 240 Mbps for all
frame
Figure 8. The Value of Throughput on the Protection Ring Network for Actual Link Plan,
Ethernet over SDH and MPLS-TP
Figure 8 indicated the throughput utilization of those techniques on the protection ring,
where the throughput utilization for Ethernet over SDH varied from 93.72 % to 93.83 % for all
frame sizes, whereas the throughput utilization for MPLS-TP is 100 % for all frame sizes.

 ISSN: 1693-6930
206
Figure 8. The Value of Throughput on the Protection Ring in Terms of Utilization in Percentage
from Actual Link Plan, Ethernet over SDH and MPLS-TP
From Figure 8 and Figure 9, we can see that there is gap between Ethernet over SDH
technique and actual link plan in term of throughput values, where the value of throughput from
the Ethernet over SDH technique didn’t reach the defined value that is 240 Mbps. Whereas
there is no gap between MPLS-TP technique and actual link plan.
The results verify that MPLS-TP with ring protection has superior performance as
compared to Ethernet over SDH with ring protection when the throughput values of MPLS-TP
with ring protection meet to 240 Mbps as a defined value. However, since throughput values of
MPLS-TP with ring protection meet to defined value, there is a greatly affect to performance of
LTE network, that is no packet loss during transmission. Conversely, when the throughput
values of Ethernet over SDH with ring protection below the defined value, therefore, it affected
to reduce performance of LTE network lead to packet loss during transmission process.
4.6. Frame Loss on the Ring Protection Network
Figure 10 shows the frame loss on the ring protection network between Ethernet over
SDH and MPLS-TP with different frame length. The results showed the frame loss values of
Ethernet over SDH varied from 14.8 Mbps to 15.08 Mbps for all frame sizes, whereas, there is
no frame loss in MPLS-TP.
Figure 10. The Value of Frame Loss on the Ring Protection Network from Actual Link Plan,
Ethernet over SDH and MPLS-TP
In Figure 11 shows the values of frame loss in percentage on the ring protection
network, where the percentage values of frame loss in Ethernet over SDH varied from 6.17 % to
6.28%, whereas, there is no frame loss in MPLS-TP. All the results showed that MPLS-TP with
ring protection provide better performance then Ethernet over SDH with ring protection.

207
Figure 11. The Value of Frame Loss in Percentage from Actual Link Plan on the Ring
Protection Network between Ethernet over SDH and MPLS-TP
4.7. Latency on the Ring Protection Network
Figure 12 illustrates the results of measurements for latency values of Ethernet over
SDH with ring protection and MPLS-TP with ring protection with different frame length. Both of
the results is kept below the standard QoS that is 150ms [16]. However, the results verify that
MPLS-TP with ring protection has better performance as compared to Ethernet over SDH with
ring protection.
Figure 12. The Latency Values on the Ring Protection Network between Ethernet over SDH and
MPLS-TP
4.8. Jitter on the Ring Protection Network
Table 5 shows the result of measurements for both Ethernet over SDH and MPLS-TP.
The results showed that the jitter’s values for both Ethernet over SDH with ring protection and
MPLS-TP with ring protection is matched with defined category that us 0-75ms [16]. The values
of latency Therefore, the value of jitter for both Ethernet over SDH with ring protection and
MPLS-TP with ring protection are available to support LTE network.
Table 5. Jitter on the Ring Protection Network
Maximum
(ms)
Maximum
(ms)
Current
(ms)
Average
(ms)
Estimate
(ms)
Category
0.015 0.015 0.015 0.015 0.015 0 – 75 ms

 ISSN: 1693-6930
208
4.9. Recovery Time
In the testing for recovery time, the data packets are routed from the main ring network
into the ring protection network due to the failure on the main ring network. Table 6 and Table 7
shows that the recovery time between Ethernet over SDH and MPLS-TP on the protection
network does not have a significant difference. Table 6 and 7 indicated that the duration to
switch back from the ring protection network to the main ring network is in accordance with the
standard ITU-T G803.2, all the values are less than 50ms. From Figure 8 to Figure 12, it can be
seen that the value of throughput, loss ratio, and jitter on the ring protection network between
Ethernet over SDH and MPLS-TP indicated that MPLS-TP has better performance compare to
Ethernet over SDH. Therefore, it can be concluded that the protection ring network is able to
maintain network performance even though the main ring network is currently unavailable.
Table 6. Recovery Time on Ethernet Over SDH
Frame Length
(Byte)
Recovery
Time (ms)
Category
64 48.097 ≤ 50 ms
128 48.320 ≤ 50 ms
256 48.590 ≤ 50 ms
512
1024
1280
1518
48.887
48.970
49.178
49.397
≤ 50 ms
≤ 50 ms
≤ 50 ms
≤ 50 ms
Table 7. Recovery Time on MPLS-TP
Frame Length
(Byte)
Recovery
Time (ms)
Category
64 48.091 ≤ 50 ms
128 48.310 ≤ 50 ms
256 48.578 ≤ 50 ms
512
1024
1280
1518
48.878
48.951
49.145
49.388
≤ 50 ms
≤ 50 ms
≤ 50 ms
≤ 50 ms
5. Conclusion
Analyzing the results, we see that throughput values of MPLS-TP with ring protection
reached 100% (240 Mbps), and throughput values of Ethernet over SDH with ring protection
varying around 93% (224.92 Mbps – 225.2 Mbps). Refer to [17] both of values are able to
support LTE network (MIMO 2 x 2) but MPLS-TP with ring protection has a better performance
than Ethernet over SDH with ring protection. The latency values of MPLS-TP with ring
protection in this paper is around 0.078ms – 0.0218ms, while the latency values of Ethernet
over SDH with ring protection is around 0.048ms – 0.1502ms, both of values are consistent with
latency standard (<150ms) [17]. From the latency values, we see that MPLS-TP with ring
protection has better performance as compared to Ethernet over SDH with ring protection.
Table 5 shows that the result of jitter measurements for both techniques are same, that is
0.015ms, this results are meet to latency standard in [17]. The last parameter that is measured
is the recovery time, which is the time needed to switch back to the main ring network, there is
no significant difference between Ethernet over SDH and MPLS-TP. The recovery time for
Ethernet over SDH is ranging from 48.097ms – 49.397ms, while recovery time for MPLS-TP is
ranging from 48.091 ms to 49.388 ms, however, it is still in accordance with the standard that
must be below 50ms. From all the testing results, we see that MPLS-TP with ring protection has
a better performance in the current assessment to satisfy LTE backhaul requirements. In terms
of link utilization, the used of MPLS-TP are meeting to the requirements of backhaul technology
that can support Industrial Internet of Things (IIoT) but need further research to improve the
overall QoS.

209
References
[1] G Kalimani Venkatesan, Kishor Kulkarni. “Wireless backhaul for LTE – requirements, challenges and
options”. in Proc. ANTS’08. 2nd International Symposium. 2008: 1-3.
[2] Ted H Szymanski. “Supporting consumer services in a deterministic industrial internet core network”.
IEEE Communications Magazine. 2016; 54: 110-117.
[3] Aleksandra Checko, Lars Ellegaard, Michael Berger. “Capacity planning for carrier ethernet LTE
backhaul networks”. in Proc. Wireless Communications and Networking Conference (WCNC). 2012:
2741-2745.
[4] Andrei Rusan, Radu Vasiu. “Assessment of packet latency on the 4G LTE S1-U interface: impact on
end-user throughput”. in Proc. Software, Telecommunications and Computer Networks (SoftCOM).
2015: 305-309
[5] Andrei Rusan, Radu Vasiu. “Emulation of backhaul packet loss on the LTE S1-U interface and impact
on end user throughput”. in Proc. Intelligent Computer Communication and Processing (ICCP). 2015:
529-536.
[6] Yi Shi, Mingchao Li, Xin Xiong, Guanglin Han, Lei Wan, Xiaodai Dong, “A Flexible Backhaul
Architecture for LTE-Advanced”. in Proc. International conference on connected vehicles and expo
(ICCVE). 2014: 1098-1105.
[7] G Vijayalakshmy, G Sivaradje. “LTE backbone for heterogeneous architectures”. in Proc.
International Conference on Circuit, Power and Computing Technologies [ICCPCT]. 2014: 1354 –
1359.
[8] Christian Addeo, Italo Busi. “Protection switching in packet transport rings”. in Proc.
Telecommunications Network Strategy and Planning Symposium (NETWORKS), XVth International.
2012: 1-6
[9] Wenjun Xie, Shanguo Huang, and Wanyi Gu. “An improved ring protection method in MPLS-TP
networks”. in Proc. Network infrastructure and digital content. 2010 2nd IEEE International
Conference. 2010: 1056-1060.
[10] Jeong-dong Ryoo, Taesik Cheung, Daniel King, Adrian Farrel, Huub van Helvoort. “MPLS-TP linear
protection for ITU-T and IETF”. IEEE Communications Magazine. 2014; 52(12): 16-21.
[11] Chang-Gyu Lim, Soo-Myung Pahk, Young-Hwa Kim. “Model of transport SDN and MPLS-TP for T-
SDN controller”. in Proc. 18th International Conference on Advanced Communication Technology
(ICACT). 2016: 522 – 526.
[12] Shicheng Zhang, Wei Ji, Xiao Li, Kangrui Huang, and Zenglu Yan. “Efficient and reliable protection
mechanism in long-reach PON”. IEEE/OSA Journal of Optical Communications and Networking.
2016; 8(1): 23 – 32.
[13] Amra Imamovic, Vlatko Lipovac. “Practial benchmark analysis of traffic protection with ethernet-over-
SDH transmission”. In Proc. Ultra Modern Telecommunications and Control Systems and Workshops
(ICUMT), 3th International Congress. 2011: 1-7.
[14] Patteti Krishna, Tipparti Anil Kumar, Kalithkar Kishan Rao. “Throughput evaluation of downlink
multiuser-MIMO OFDM-LTE System”. Int. J. Communications, Network and System Sciences. 2015;
8: 58-61.
[15] IXIA. “Predicting and managing network impairments”. White paper, rev. A. 2012.
[16] Nigel Burgess. “RFC 2544 Testing of Ethernet Services in Telecom Networks”. White Paper, Agilent
Technologies. 2004.
[17] NEC, “SDH Transmission System”. 6
th
ed, Japan. 1999.

Capacity Improvement and Protection of LTE Network on Ethernet Based Technique

More Related Content

What's hot (20)

Similar to Capacity Improvement and Protection of LTE Network on Ethernet Based Technique (20)

More from TELKOMNIKA JOURNAL (20)

Recently uploaded (20)

Capacity Improvement and Protection of LTE Network on Ethernet Based Technique