Performance analysis of voip traffic over integrating wireless lan and wan using different codecs

International Journal of Wireless & Mobile Networks (IJWMN) Vol. 6, No. 3, June 2014
DOI : 10.5121/ijwmn.2014.6306 79
PERFORMANCE ANALYSIS OF VOIP TRAFFIC OVER
INTEGRATING WIRELESS LAN AND WAN USING
DIFFERENT CODECS
Ali M. Alsahlany1
1
Department of Communication Engineering, Al-Najaf Technical College, Foundation
of Technical Education, Iraq
ABSTRACT
A simulation model is presented to analyze and evaluate the performance of VoIP based integrated
wireless LAN/WAN with taking into account various voice encoding schemes. The network model was
simulated using OPNET Modeler software. Different parameters that indicate the QoS like MOS, jitter,
end to end delay, traffic send and traffic received are calculated and analyzed in Wireless LAN/WAN
scenarios. Depending on this evaluation, Selection codecs G.729A consider the best choice for VoIP.
KEYWORDS
VoIP, Codecs, QoS
1. INTRODUCTION
During the recent years, there is a growing trend in real time voice communication using
Internet protocol (IP). Voice over Internet Protocol (VoIP) is a technology that allows delivery
of voice communications over the Internet or other packet switched networks rather than the
traditional Public Switched Telephone Network (PSTN). Many VoIP applications are available
on the internet: Skype, Viber, Tango, and Yahoo messenger. All of these applications provide
good quality, and free calls. In VoIP the analogue voice signal of the transmitter is converted
into digital format before compression and encoding it into a stream of IP packets for
transmission to the receiver over IP network. At the receiving end, Digital to Analogue
Converter (DAC) works on regenerating the original analogue voice signal after reassemble
received IP packets in order and processing it [1].
There are many authors have worked on various Quality of Service (QoS) parameters using
different service classes in different network types. A study was conducted on various quality
parameters impacting on the VOIP service performance. The study refers that these parameters
of QoS are required to increase the performance of a VoIP. Researchers in [2], have compared
the performance of the VoIP in both Ethernet LAN (802.3) and Wireless LAN (IEEE 802.11).
They examine how VoIP performs in two different network setups and analyzes the results
obtained using OPNETsimulator. They also examine the optimization of IEEE 802.11e for QoS
using the priorities to provide real time service for VoIP. Various QoS parameters like
throughput and average delay for VoIP using different protocols are analyzed in Ref. [3] using
OPNET simulator. Simulation results show that the Optimized Link State Routing protocol has
better performance in terms of throughput and average delay. Similar analysis has been
conducted in [4] to analyze the QoS of VoIP deployment over WiMAX network and compare

80
the performance obtained over various service classes. In [5] authors discussed a handoff
mechanism effect on VoIP traffic in Wireless LANs for intra and inter mobility. The study was
oriented towards the assessment of the quality of the voice traffic, the packet delays, and delay
jitter during the handover operation. Atiur et.al. [6], used an analytical model to estimate the
VoIP call capacity of IEEE 802.11g wireless LAN to support single hop networks and multi-
hop VoIP services. The authors discovered that, the number of hops between VoIP transmitters
in IEEE 802.11g wireless LAN has effects on call quality. Hussein et al. [7], conducted a
comparison between different queue algorithms. The authors found that Priority queue and
Weight Fair Queuing algorithms are the most appropriate to improve QoS for VoIP. In [8]
authors used VoIP as the multimedia measurement to identify the main limitations for
enhancing QoS in Wireless LANs, compared to wired networks.
However, none of them have presented accurate approximations to predict the performance of
QoS for VoIP when various audio codec schemes are used. Also, most of these studies were
performed in a small distance network area without taking real WAN as a one of real
application environment. In his paper, the OPNET simulator used to implement the proposed
VoIP Network. The simulation model for the proposed network is a private network for a
company has two locations located at two different countries around the world in order to
simulate the communications between two locations as a long distance and the same location as
a local. The major factors that affect on the QoS for VoIP such as: mean opinion score (MOS),
average end to end delay, jitter, traffic send, and traffic received are analyzed according to
International Telecommunication Union (ITU) standards.
A comparison was carried out between different codecs (G.711, G.729A and G.723.1) which
are the most appropriate to improve QoS for VoIP.
The rest of this paper is organized as follows. Section 2 focus on VoIP codec schemes. The
model design and configuration networks are described in Section 3. Section 4 presents
simulation results and analysis. Finally, Section 5 concludes the paper.
2. VOIP CODECS
A codec is the term used for the word coder-decoder, converts of analog audio signals into
compressed digital form for transmission and then back into an uncompressed audio signal
for the reception. There are different codec types based on the selected sampling rate, data
rate, and implemented compression. the most common codecs used for VoIP applications
are G.711, G722, G723, G726, G728, G729A, etc. each of which varies in the sound
quality, the bandwidth required, the computational requirements, encoding algorithm and
coding delay [3, 9, 10]:
In our evaluations codecs of ITU standards for audio compression and decompression are
used. Following table lists some features of the most common codecs: G.711, G.723.1, and
G.729A.

81
Table 1: Features of the most common codecs: G.711, G.723.1 and G.729A
IUT-T
Codec
Algorithm
Codec
Delay
(ms)
Bit Rate
(Kbps)
Packets
Per
Second
IP Packet
Size
(bytes)
Comments
G.711 pulse-code
modulation (PCM)
0 .375 64 100 120
Delivers precise speech
transmission. Very low
processor requirements
G.723.1
Multipulse
maximum likelihood
quantization/algebra
c-code excited linear
prediction
97.5 5.3 33 60
High compression with
high quality audio. Can
use with dial-up. Lot of
processor power
G.729A
Conjugate Structure
– Algebraic Code
Excited Linear
Prediction (ACELP)
35 8 100 50
Excellent bandwidth
utilization. Error
tolerant
3. NETWORK MODEL DESCRIPTION
Our simulation approach uses OPNET simulator for network modeling. OPNET simulator is an
authoritative communication system simulator developed by OPNET Technologies.
This simulation model was run in three different scenarios to evaluate the difference in
performance and to determine the best audio encoding schemes of utilizing VoIP over
integrating Wireless LAN/WAN. All the scenarios follows the same structure and the same
topology. The basic difference in all the scenarios is with the configuration of the corresponding
application and the profiles. As one scenario is implemented with the codec G.711, G.723.1,
and G.729.1. Various comparisons conducted to find the values of various parameters.
The simulation model for the network under study is illustrated in Figure 1. The Application
Config, Profile Config, Mobility Config and QoS Attribute Config are included in the model.
The Wireless LANs has been considered as the components of the whole network and its nodes
are using VoIP services.
A frame relay model supporting Permanent Virtual Circuit (PVC) implemented for the WAN.
At each end of the frame relay we had subnets. Inside these subnets was a Wireless LAN
environment. The Wireless LAN environments consisted of 3 Routers and 5 workstations in
each Router. All the Routers connected together using a Switch. At wireless subnet_0 as shown
in Figure 2, Two Servers were connected to the Ethernet Router titled VoIP_Server and Server.
The VoIP_Server handled the VOIP traffic and the Server handled the FTP and HTTP traffic.
All The Wireless LAN Routers were connected to the frame relay via an Ethernet Router that
connected the Switch. The WAN and Wireless LAN were both star topologies.

82
Figure 1: The Simulation Network Model
Figure 2: The wireless subnet_0

83
Figure 3: The wireless subnet_1
3.1 BASIC PARAMETER
The Point to Point WAN was connected directly to the Ethernet Router in each subnet using a
DS0 line. The model name of the link was ‘point_to_point_link_adv’. The Wireless LAN
environment used 1Gbps Ethernet cable for all the wired connections to the Router, Switch, and
Servers. The OPNET model name for the Servers was ‘ethernet_server’. Each Wireless LAN
environment contains a Router and 2 fixed wireless workstations and 3 moving wireless
workstations in each Router. All these wireless nodes use VoIP services. The Routers and
workstations were using 802.11g running at a data rate of 54 Mbps. There were a total of 15
workstations in each subnet. The workstations were generating traffic across the WAN and the
Wireless LAN environment to simulate a real office environment. The Routers OPNET model
name was ‘wlan_ethernet_slip4_adv. The fixed VoIP workstation model and the moving VoIP
workstation model name were ‘wlan_wksn_adv’.
All the parameters using in the Wireless LAN environment have the same value in the same
subnet. The Wireless LAN attributes are shown in Table 2.
Table 2: Parameters used in simulation
Parameter Value
Physical characteristic IEEE 802.11g
Data Rate 54 Mbps
Roaming Capability Enable
Transmit Power 0.005 W
Power Threshold -95 db
Short Retry Limit 7
Long Retry Limit 4

84
3.2 APPLICATION, PROFILE, AND MOBILITY CONFIGURATIONS
Application Definitions were where the applications were created for traffic generation on the
network. There are 3 applications used in each scenario. These are VoIP, FTP and web
browsing applications. The important one was VoIP. In the Voice category IP Telephony
selected for configuration VoIP. In the VoIP application, voice attributes setup as codec scheme
G.711 for scenario 1; G.723.1 for scenario 2; and G.729A for scenario 3. G.729A operates at a
bit rate of 8 Kbps, G.723.1 operates at a bit rate of 5.3/6.3 Kbps and G.711 operates at a bit rate
of 64 Kbps. For the QoS scenarios the VoIP table type of service (ToS) was changed to 8 to
give voice priority. Other parameters that were set for this project were file transfer and web
browsing. For each of these applications mentioned there was a heavy usage parameter set for
each one.
Profile Configuration describes the activity applications used by users through a period of time.
The FTP_Profile was using VoIP, FTP, Web Browsing and File Transfer with light usage
parameters. VoIP_Profile was only using VoIP applications, these were the dedicated IP
phones. All the Profiles configured to run simultaneously to allow more than one application to
operate at the same time.
The Profile Definition for VoIP traffic is configured that the first VoIP call is starting at 100
second (60 seconds offset time + 40 seconds start time) of the simulation, after that one VoIP
call is added after 1 second. The addition of VoIP calls is prepared by repeating the voice
application in the Profile Definition. This procedure continues till the end of the simulation.
Mobility Configuration setup the trajectory for the wireless mobile nodes in the Wireless LAN
environments. There were two vector mobility profiles. The differences in the profiles were
mobility speed.
4. SIMULATION RESULTS AND ANALYSIS
The following figures are obtained after collecting statistics by using OPNET Modeler
simulation tool. Each figure shows a comparative picture of the three scenarios. All the three
scenarios are using a different audio codec scheme such as G.911, G.923 and G.929. After
successfully running the simulation, The result shows the impact of different codecs on different
QoS parameters in a VoIP network. Following are the figures that show different QoS
parameters like MOS, Voice packet end to end delay (sec), Voice jitter (sec), Voice traffic
sends (packet/sec) and Voice traffic received (packet/sec).
The most widely used QoS metric in VoIP applications is MOS. The MOS value describes the
voice perception quality. The average MOS value for the three codecs is represented in Figure
4. Codecs G. 711 and G. 729A have acceptable MOS values 3.685 and 3.067, respectively. On
the other hand, the MOS value for G. 723.1 is 2.557 which indicates that the quality of service
is poor if this codec used.

85
Figure 4: Average voice MOS under various codecs
Average end to end delay metric is shown in Figure 5. G.729A presents the best performance
with respect to other codecs. These results are due to transfer rate and packet size. The low
packets transfer and the larger packet size, the more time is required to process them. The
relatively high transfer rate (8 kbps) and low packet size (20 bytes) for G.729A make G.729A
the ideal codecs. Otherwise, G.723.1 and G.711 suffered highest delay than G.729A for the
reason that it has the lowest bit rate (5.3 kbps for G.723.1) and larger packet size (160 bytes for
G.711). In turn, the end to end delay is increased with transfer rate and packet size.

86
Figure5: Average voice packet End to End delay (sec) under various audio codecs
Figure 6 describes the average voice jitter comparison using different codecs. From the Figure,
the variation of the codec G.729A is minimum and approximately constant throughout the
simulation. The average voice jitter variation in case of codec G.723.1 is higher than the other
two codecs at the earlier time of simulation, but after some time it falls down. The jitte
variation in case of G.711 lies between two other audio codecs. The voice jitter threshold for
smooth communication in VoIP network is about 1 ms [11] so audio codec G.729A gives better
results than audio codecs G.711 and G.723.1 respectively. So there is a high increase in jitter as
audio codecs G.711 and G.723.1 are added to the network. This increase in voice jitter makes
the voice difficult to understand due to arriving packets at different time . The use audio codec
G.729A will make the jitter less and best performance of VoIP application in Integrating
Wireless LAN and WAN.

87
Figure 6: Average Jitter under various audio codecs
Average voice traffic sent and received is presented in Figure 7. Any network to be more
efficiency these two traffics must be equal. The traffic received by the network with codec G.
729 is less deviated from the traffic sent comparatively with codec G.723.1 and G.711. This
analysis indicates that the noise added in the G.729A network is less when compared to the
other networks, so this codec is more efficient.

88
Figure 7: Average voice traffic sent and received under various audio Codecs.
5. CONCLUSION
In this paper, analysis and evaluation of the QoS performance for VoIP traffic under various
voice codecs was carried out. The use of codecs appropriately is very important in the
implementation of VoIP to generate maximum QoS value. The result shows a selection of
G.729A codec in a simulation gives a significant result for the performance of VoIP that codec
G.729A has acceptable MOS value and less deviation of received to transmit packet as
compared to G.711 and G.723.1 also average delay like end to end delay and Voice jitter is
lesser in codec G.729A as compared to the other two referenced codecs.

89
REFERENCES
[1] Y. Jung, and C. Manzano, "Burst packet loss and enhanced packet loss-based quality model for
mobile voice-over Internet protocol applications," Journal of IET Commun., Vol. 8, Iss. 1, pp. 41–49,
2014.
[2] K. Neupane, and et al, " Measuring the Performance of VoIP over Wireless LAN," In Proceedings of
the 2011 conference on Information technology education (SIGITE '11), New York, USA, pp. 269-
274, 2011.
[3] S. Anand, and et al, " Performance Analysis of VoIP Traffic using various Protocols and Throughput
enhancement in WLANs," International Conference on Computer, Communication and Electrical
Technology," March 18-19, pp. 176-180, 2011.
[4] T. Anouari, and A. Haqiq, " Performance Analysis of VoIP Traffic in WiMAX using various Service
Classes," International Journal of Computer Applications,Vol. 52, No.20, August, pp. 29-34, 2012.
[5] Abderrahmane Lakas, " Experimental Analysis of VoIP over Wireless Local Area Networks,"
Journal of Communications, Vol. 2, No. 4, June , pp. 3-9, 2007.
[6] M. Atiur, and et al, " An Analytical Approach for Voice Capacity Estimation Over WiFi Network
Using ITU-T E-Model," IEEE Transactions on Multimedia, Vol. 16, No. 2, pp. 360-372, Feb., 2014.
[7] Hussein, and et al, " The Effects of Different Queuing Algorithmswithin the Router on QoS VoIP
application Using OPNET," International Journal of Computer Networks & Communications (IJCNC)
Vol.5, No.1, Jan , pp. 117-124, 2013.
[8] Ayman Wazwaz, and et al, " Analysis of QoS parameters of VOIP calls over Wireless Local Area
Networks" The 13th International Arab Conference on Information Technology ACIT', Dec.10-13,
pp. 409-414, 2012.
[9] T. Daengsi, and et al, "A Study of Perceptual VoIP Quality Evaluation with Thai Users and Codec
Selection Using Voice Quality - Bandwidth Tradeoff Analysis," ICT Convergence (ICTC), 2013
International Conference ,pp.691-696, 2013.
[10] S. Brak , and et al, " Speech Quality Evaluation Based CODEC for VOIP over 802.11P,"
International Journal of Wireless & Mobile Networks (IJWMN), Vol. 5, No. 2, April, pp. 59-69,
2013.
[11] U. R. Alo, and Nweke Henry, " Investigating the Performance of VOIP over WLAN in Campus
Network," Journal of Computer Engineering and Intelligent Systems, Vol.4, No.4, pp. 47-58, 2013.
Author
Ali M. Alsahlany (alialsahlany@yahoo.com) has got his MSc in communications
engineering from Basrah University in 2012. He is now working as a lecturer for
computer networks and wireless communications engineering in Communication
Engineering Department at Foundation of Technical Education. His area of interest
includes Computer Networks, Wireless and Mobile Networks, Optical
Communicationand , and Data Communications.

Performance analysis of voip traffic over integrating wireless lan and wan using different codecs

More Related Content

What's hot (17)

Viewers also liked (19)

Similar to Performance analysis of voip traffic over integrating wireless lan and wan using different codecs (20)

Recently uploaded (20)

Performance analysis of voip traffic over integrating wireless lan and wan using different codecs