International Journal of Computer Applications (0975 8887)

Volume 84 No 5, December 2013

Selection of VoIP CODECs for Different Networks based

on QoS Analysis
Malik Ahsan Ali

Imran Rashid

Adnan Ahmed Khan

Electrical Engineering
Department
MCS, National University of
Science and Technology,
Islamabad, Pakistan.

Electrical Engineering
Department
MCS, National University of
Science and Technology,
Islamabad, Pakistan.

Electrical Engineering
Department
MCS, National University of
Science and Technology,
Islamabad, Pakistan.

ABSTRACT
Voice over Internet Protocol (VoIP) has been an interesting
topic of research in the last decade. The engrossing increase in
the use of VoIP services is resulting in the enormous growth
of broadband network. The main objective of this paper is the
selection of an appropriate voice compression and
decompression (CODEC) schemes depending on the Quality
of Service (QoS) of VoIP in different networks. Wired,
Wireless Local Area Network (WLAN), Worldwide
Interoperability for Microwave Access (WiMAX) and
Universal Mobile Telecommunication System (UMTS)
networks were implemented in OPNET Modeler. The quality
is compared using different QoS parameters like end-to-end
delay, MOS, throughput and jitter. The VoIP codecs used in
the measurements of QoS are: GSM-FR, G.711, G.723.1 and
G.729A. Simulations showed that G.711 and GSM- FR are
the best schemes that provide high quality of voice in
Wireless Local Area Network (WLAN) communications. In
WiMAX, G.729A gives the best quality of VoIP while in
UMTS, GSM- FR gives overall best results with respect to all
the parameters. Wired model gives the best result irrespective
of the codec being used. G.723.1 can be used in WiMAX and
UMTS along with the wired network depending on
conditions. The results analyzed and the performance
evaluated will give network operators an opportunity to select
the codec for better services of VoIP for customer satisfaction.

Keywords
VoIP; WLAN; WIMAX; UMTS; Codec; QoS.

1. INTRODUCTION
Voice over Internet Protocol (VoIP) practices is potentially
mounting day by day resulting in the demand of rapid
improvements in the networks. There is a demand of
decreasing the difference between the qualities of voice and
increasing the available bandwidth to provide the best VoIP
services comparative to the traditional circuit switched
telephony [1]. VoIP has almost replaced the conventional
Public Switched Telephone Network (PSTN) due to its cost
effectiveness and the features being provided [2]. The wired
Internet Protocol (IP) networks provide better VoIP services
as compared to the wireless IP network as wireless networks
have their own characteristics and impairments [3]. The
unsolved issues caused by the wireless network in this area
still needs some dedicated work spotlighting VoIP calls. In
next generation networks wired and wireless systems have
been combined in an innovative way under a single
framework [5]. The frequent handovers cause delay and
packet loss in these network [6]. The VoIP call gets degraded
and loses the packets more swiftly. An eternal solution is
required for these heterogeneous systems for the VoIP
communication.

The major benefits that wireless network provide is the

mobility of the users. It becomes even more challenging on
combining the wireless network with multimedia and realtime services to provide QoS [7]. In wireless communication
fading influences the channel and degrades the quality of
VoIP [8]. WiMAX network has remained the main area of
interest for the researchers for VoIP analysis and other
networks have rarely been studied [9]-[12]. A correlation has
been concluded between the codec being used, amount of
voice frames and the delay [13]. The effect of delay on VoIP
calls over the WiMAX network depends on the distance and
number of base stations [14]. The performance of G.723.1
VoIP CODEC and measurement of the capacity of a WiMAX
test bed for up-link and down-link was studied in [15]. Mean
Optimum Score (MOS) is the main parameter on which the
quality of voice can be concluded. The performance of
network using Session Initiation Protocol (SIP) is compared in
IPv4 and IPv6 networks [16]. The Windows Operating
systems can use either IPv4 or IPv6 and to obtain best VoIP
performance the preferred codec can be selected [17].
The researchers have been able to provide some good results
but as most of the IP networks and their underlying protocols
[4] in use today were implemented keeping in mind the data
services not the real-time and delay sensitive voice services
[18], there is still a need of providing better QoS as per the
demand of the users. A variety of CODECs are being used
today each having its own characteristics [19]. Mostly used
codec for VoIP is G.711 which provide good results for wired
network depending on the environment and conditions [20]
but when it comes to wireless networks the quality is
degraded. The internet services are becoming so complex that
VoIP performance parameters require some actual
measurements unlike the traditional telephone networks which
were dependent on mathematical modeling. This research is to
provide good quality of VoIP services in every network and
analysis is done using different codecs mentioned in table1.
VoIP packets are analyzed focusing all the major parameters
like end-to-end delay, MOS, throughput and jitter over Wired,
Wireless, UMTS and WiMAX networks using the OPNET
Modeler..

2. VOIP AND CODECS

The demand for mobile and broadband services is rising day
by day. The last decade has seen the ever-increasing VoIP
users with the demand of reliable and quality services. VoIP
is an emerging technology for voice communication used
these days. The services are not only being used for long
distance calls but also for the short distant communications.
The devices like IP phones and the VoIP enabled desktop
systems are cost effective and also provide some new features
to the users. Keeping in mind the demand of the users, the
operators are forced to improve the quality of communication.

38

International Journal of Computer Applications (0975 8887)

Volume 84 No 5, December 2013
This can be achieved by increasing the bandwidth and making
the IP backhaul that fulfills the demand of the users at lower
cost providing better QoS.

Fig. 1: Wired Model

3.2 WLAN(wireless)

2.1 VoIP Codecs

Codec is a coder/decoder which converts the audio signal to
digitized version for transmission over the medium and then
back into the original uncompressed version on the receiver
side. This concept is the base of VoIP services. There are a
number of codec used for VoIP communication each having
its own bandwidth and characteristics. The codecs which are
used in this research work are listed in the table I below.
Table 1. Characteristics of VoIP Codecs
CODEC
GSM- FR
G.711
G.723.1

Coding Algo
PRE-LTP
PCM
ACELP

Sampling rate
13 kbps
64 kbps
5.3 kbps

G.729A

CS-ACELP

8 kbps

Fig. 2: WLAN Model

3.3 UMTS

3.

NETWORK MODELS

The tool used for simulations is OPNET Modeler as it

provides the results very closer to the real time environment.
The models were created by selecting the nodes and links
from the object palette such that to reduce the
losses/impairments effect. Wired model designed, is a general
IP network. Links in the wired design as shown in figure 1
consist of standard 100baseT lines from user to router and
from router to internet cloud followed internet server is T1
line. WLAN design consists of user node and access point
connected to the IP backhaul with a T1 line as shown in figure
2. UMTS model as in figure 3 comprises user equipments,
node B and Radio Network Controller (RNC) which is
connected to the packet switched network via Serving GPRS
Support Node (SGSN) and Gateway GPRS Support Node
(GGSN) which in turn is connected to the IP Network. Figure
4 represents the WiMAX model which is designed using the
base station connected to the IP backhaul serving the VoIP
users. A T1 line is used to simulate a perfect connection
between router and server minimizing cable delay and
allowing the difference caused by the codecs to be more
noticeable. The attributes and parameter settings are made in
the network models and various simulations are carried out
for the codecs. The reason for utilizing this modeling method
is to allow performance of the codecs to be analyzed in an
improved manner.

3.1 Wired

Fig. 3: UMTS Model

3.4 WiMAX

Fig. 4: Wimax Model

39

International Journal of Computer Applications (0975 8887)

Volume 84 No 5, December 2013

4. RESULTS AND ANALYSIS

The comparative analysis of UMTS, WiMAX, Wired and
WLAN networks using each codec while keeping the
simulation environment and attributes same, is discussed in
this section. The performance of each codec is evaluated in
the network models depending on the QoS.

4.1 Analysis of Codec G.711

This simulation is performed for G.711 codec in different
networks. The results shown below are used to evaluate the
performance of G.711. It is analyzed from figure 5 that the
value of Mean Opinion Score (MOS) is 3.7 in wired, wireless
and WiMAX models, showing the good quality of speech.
Comparatively, MOS value for UMTS model is 2.5 showing
the worst quality of speech amongst all the models. Jitter and
end to end delay in figures 6 and 7 shows that UMTS and
WiMAX models undergo a delay in packets and attain some
jitter. In UMTS a jitter and significant amount of delay is
attained degrading the quality while in WiMAX, delay and
packet loss effect the communication. Traffic sent is almost
same in all the models as shown in figure 8 while the traffic
received in figure 9 shows that in WiMAX and UMTS there is
loss in the packets as compared to wired and WLAN
networks. Jitter, delay and less reception of packets in UMTS
model represent that it gives worst quality of voice while
using G.711. The performance of WiMAX models is also not
effective as there is a delay and it loses packets. Wired and
WLAN models give best performance while using G.711.

Fig. 6: Jitter for G.711

Fig. 7: Packet end to end delay for G.711

Fig. 5: MOS for G.711

Fig. 8: Traffic sent for G.711

40

International Journal of Computer Applications (0975 8887)

Volume 84 No 5, December 2013

Fig. 9: Traffic Received for G.711

Fig. 11: Jitter for G.723.1

4.2 Analysis of Codec G.723.1

This simulation is performed for G.723.1 codec in different
networks. The results shown in figure 10-14 are used to
evaluate the performance of G.723.1. Figure 10 show that the
value of MOS is 2.5 in wired, WLAN and WiMAX models.
Comparatively, MOS value for UMTS model is 2 showing the
bad quality of speech. However when it comes to jitter and
end to end delay in figures 11 and 12, WLAN and UMTS
models along with the wired model have minimum delay in
packets and attain zero jitter. In WiMAX jitter and the amount
of delay is very small hence providing the good quality of
VoIP. Traffic sent and received is almost same in all the
models except WiMAX model which loses a small amount of
packets as shown in figures 13 and 14. Jitter, delay and full
reception of packets in wired, WLAN and UMTS models
allow using G.723.1 codec with low MOS. The performance
of wired, WLAN, WiMAX and UMTS models is not quite
effective but the jitter and delay is reduced using G.723.1
codec.
Fig. 12: Packet end to end delay for G.723.1

Fig. 10: MOS for G.723.1

Fig. 13: Traffic sent for G.723.1

41

International Journal of Computer Applications (0975 8887)

Volume 84 No 5, December 2013

Fig. 14: Traffic Received for G.723.1

Fig. 16: Jitter for.729A

4.3 Analysis of Codec G.729A

This simulation is performed for G.729A codec in different
networks. The results shown in figure 15-19 are used to
evaluate the performance of G.729A. Figure 15 show that the
value of MOS is 3.2 in wired, WLAN and WimAX models.
Comparatively, MOS value for UMTS model is 2.4 showing
the bad quality of speech. WiMAX model along with the
wired model show the best quality of VoIP. Jitter and end to
end delay in figures 16 and 17 shows that WLAN and UMTS
models undergo a delay in packets and attain some jitter
which in turn loses the packets. In WiMAX jitter and the
amount of delay is very small hence providing the good
quality of VoIP. Traffic sent and received is almost same in
wired and WiMAX models while there is some loss of packets
in WLAN and UMTS networks as shown in figure 23 and 24.
Jitter, delay and full reception of packets in WiMAX model
represent that it gives best quality of voice while using
G.729A. The performance of WLAN and UMTS models is
not effective as there is a delay and packets are lost.
Fig. 17: Packet end to end delay for G.729A

Fig. 15: MOS for G.729A

Fig. 18: Traffic sent for G.729A

42

International Journal of Computer Applications (0975 8887)

Volume 84 No 5, December 2013

Fig. 19: Traffic Received in G.729A

Fig. 21: Jitter for GSM-FR

4.4 Analysis of Codec GSM-FR

This simulation is performed for GSM-FR codec in different
networks. The results shown below in figure 20-24 are used to
evaluate the performance of GSM- FR. Figure 20 shows that
the value of MOS is 3.6 in wired and WLAN models.
Comparatively, MOS value for UMTS model is 3 showing the
good quality of speech. WiMAX model shows the worst
quality amongst all the networks. Jitter and end to end delay
in figures 21 and 22 shows that WiMAX model undergoes a
delay in packets and attain some jitter which in turn loses the
packets. In UMTS jitter and the amount of delay is very small
hence providing the good quality of VoIP. Traffic sent and
received is almost same in all the models except WiMAX as
shown in figure 23 and 24. Jitter, delay and less reception of
packets in WiMAX model represent that it gives worst quality
of voice while using GSM-FR. The performance of UMTS
model is very much effective as there is no delay and it does
not lose packets. Wired and WLAN models give the best
performance.
Fig. 22: Packet end to end delay for GSM-FR

Fig. 20: MOS for GSM-FR

Fig. 23: Traffic sent for GSM-FR

43

International Journal of Computer Applications (0975 8887)

Volume 84 No 5, December 2013
[7] Muhammad Imran Tariq, Muhammad Ajmal Azadand
Syed Khurram Rizvi. Effect of Mobility Patterns on
VoIP QoS in Mobile WiMAX, International Journal of
Computer Science and Telecommunications,vol. 4, 2013
[8] H. Arora and H. Sethu, A simulation study of the impact
of mobility on performance in mobile ad hoc networks,
in Proceedings of the Applied Telecommunication
Symposium San Diego, California, April 14-18, 2002.
[9] A. A. Ali, S. Vassilaras and K. Ntagkounakis, A
Comparative Study of Bandwidth Requirements of VoIP
Codecs over WiMAX Access Networks, Third
International Conference on Next Generation Mobile
Applications, Services and Technologies (NGMAST
'09), 2009, pp. 197 - 203.
[10] K. A. Shuaib, A Performance Evaluation Study of
WiMAX Using Qualnet, Proceedings of The World
Congress on Engineering 2009, pp. 912-916.
Fig. 24: Traffic Received for GSM-FR

5. CONCLUSION AND FUTURE WORK

Performance of various VoIP codecs in different networks is
analyzed using the OPNET Modeler. A variety of simulations
are carried out to get the most effective and efficient results.
On the basis of results attained, conclusion for the selection of
VoIP codecs in different networks is made. Depending on the
results it is concluded that wired network performs well
irrespective of the VoIP codec being used. G.711 and GSMFR can be selected for VoIP communications in WLAN
network. For WiMAX network, G.729A codec is the most
effective one. In UMTS network the best VoIP quality is
given while using GSM-FR. The quality of G.723.1 codec is
observed low as it is a low quality codec. Hence it can be used
in all the networks depending on the environment and users
density. The conclusions will be helpful for the network
operators and also for the researchers to further work on the
topic. The mobility parameter of the wireless models will be
the main focus of the future work.

6. REFERENCES
[1] T. Kwok, "Residential broadband Internet services and
applications requirements" Communications Magazine,
IEEE Volume 35, Issue 6, June 1997 Page(s):76 - 83
[2] T. Wallingford, "Switching to VoIP", Publisher:
O'Reilly, ISBN: 0-596-00868-6, Pub Date: June 2005
[3] A.
Samukic,
UMTS
Universal
Mobile
Telecommunications System Development of Standards
for the Third Generation, 1998
[4]

IEEE Std 802.11, Wireless LAN Medium Access

Control(MAC)
and
Physical
Layer
(PHY)
Specifications, ANSI/IEEE Std 802.11, 1999.

[5] Per Beming, Lars Frid, Gran Hall, Peter Malm, Thomas
Noren, Magnus Olsson and Gran Rune, LTE-SAE
architecture and performance, Ericsson Review No. 3,
2007
[6] Naveed Iqbal, Ajmal Khan, Malik Ahsan Ali, Uzma
Anwar, Burhan Ullah, M. Faizan Sabir, Performance
Analysis of Soft and Hard Handovers based on UMTS
QoS Traffic Classes, ICIIT, 2010.

IJCATM : www.ijcaonline.org

[11] S.
Jadhav,
Z.
Haibo
and
H.
Zhiyi,
PerformanceEvaluation of Quality of VoIP in WiMAX
and UMTS, 12th International Conference on Parallel
and Distributed Computing, Applications and
Technologies (PDCAT), 2011, pp. 375380.
[12] M. I. Tariq, M. A. Azad, R. Beuran and Y. Shinoda,
Performance Analysis of VoIP Codecs over BE
WiMAX Network, 3rd International Conference on
Computer and Electrical Engineering (ICCEE), 2010,
vol. 9, pp. 4751.
[13] C. Jianguo and M. Gregory, Performance Evaluation
of VoIP Services using Different CODECs over a UMTS
Network, Australasian Telecommunication Networks
and Applications Conference (ATNAC), 2008, pp. 6771.
[14] Safak and B. Preveze, Analysis of delay factors for
voice over WiMAX, in Computer and Information
Sciences, 2008. ISCIS 08.
23rd International
Symposium on, oct. 2008, pp. 16.
[15] K. Pentikousis, E. Piri, J. Pinola, F. Fitzek, T. Nissila,
and I. Harjula, Empirical evaluation of voip aggregation
over a fixed WiMAX testbed, in Proceedings of the 4th
International Conference on Testbeds and research
infrastructures for the development of networks &
communities, ser. TridentCom08, 2008, pp. 19:119:10.
[16] T. Hoeher, M. Petraschek, S. Tomic, and M.
Hirschbichler. Evaluating Performance Characteristics
of SIP over IPv6, IEEE Journal of Networks,vol. 2, no.
4, pp. 10, 2007.
[17] Hira Sathu and Mohib A. Shah. Performance
Comparison of VoIP Codecs on Multiple Operating
Systems using IPv4 and IPv6 International Journal of eEducation, Vol. 2, No. 2, April 2012
[18] R. Dominach Quality of service for
applications, TMCnet. March 2004.

real-time IP

[19] ITU-T P.830, Subjective performance assessment of

telephone-band and wideband digital codecs, 1996.
[20] Aamir, M.; Zaidi, S.M.A., "QoS analysis of VoIP traffic
for different codecs and frame counts per packet in
multimedia environment using OPNET," INMIC, 2012
15th International, vol.,no., pp.275,281, 13-15Dec.

44