Initially Internet was designed for non real time applications such as data transfer. Rapid growth of Internet, new application developments, high bandwidth demand and user requirements is becoming challenging job to handle. Networking at enterprise level in real scenario is not something, which can be done without proper investigation and planning. Technology advancement introduced different applications and protocols to data networks such as VoIP and video streaming. VoIP (Voice over IP) is a protocol, which carries voice signals over IP networks. It is a cost saving option to communicate in comparison to traditional telephony services especially for long distance calls. Most common examples are voice chat computer to computer, computer to phone etc. The purpose of this dissertation is to apply different QoS mechanisms into VoIP network and simulate impact of background traffic on QoS. Simulation based on a medium enterprise level of network will be used for comparative studies and performance evaluation.
The aim of this project is to study the effects of different QoS mechanisms on exiting VoIP network and the impact of background traffic on QoS (Quality of Service). OPNET simulation software will be used for simulation of medium enterprise level of network with 10 Routers and 200 users. Objectives
- To evaluate the performance of a medium scale VoIP network with introduction of different QoS mechanisms, i-e FIFO (First In First Out), WFQ (Weighted Fair Queue) and CQ (Custom Queuing).
- To evaluate the impact of background traffic on QoS. Outcomes
By execution of simulation model with three different QoS mechanisms alterations, data such as bandwidth utilization, throughput, jitter and end-to-end delay will be captured in numerical and graphical form for analysis and comparison. This captured data or results might be an indication of the effects of QoS mechanisms on medium enterprise level of network and the impact of background traffic on QoS.
VoIP stand for voice over IP is self-explanatory, that is carrying voice over IP (Internet Protocol) . In the field of computer networks and Internet world, VoIP is becoming very popular and expanding rapidly to replace the traditional PSTN (Public Switched Telephone Network). It is relatively cheaper in comparison to PSTN, more and more VoIP based technology devices are being developed and also video streaming is getting common. Conventional voice networks based on circuit switching requires the PBX (Private Branch Exchanges) to be reconfigured each time when relocation of staff occurs or when new staff is added to the existing infrastructure of the organization. In contrast to PBX, VoIP works like plug & play devices of an operating system. DHCP (Dynamic Host Configuration Protocol) helps to register the IP phone automatically. IP phone have more facilities then old traditional telephone system. It is flexible and with extra-added features such as to work as intercom at the same time, speed dialing, call park etc. Administrator can manage user accounts and requests to troubleshoot a problem remotely. Long distance calls are becoming cheaper, revenue and efficiency is increasing due to VoIP technology and companies are migrating to VoIP from legacy telephony system . Bandwidth requirement is less. By implementing QoS (Quality of Service), VoIP and data networks can be configured in more prominent way to get the optimum result. Important factor to consider is that how to isolate the real time traffic from data network so that it can be prioritized to achieve acceptable quality of service for VoIP traffic. Traveling, office expenses, parking problem, traffic congestion and branch offices located in different geographical locations all these factors made a base for teleworker concept. To avoid these problems more and more companies are adopting the teleworker option for their employees, that are not only cost saving for companies but also less stressful for employees in the busy life schedule of present time and improves their efficiency . Companies are migrating from legacy telephony system to VoIP technology due to cost effectiveness and flexible solution as mentioned above. However question arises that how can QoS be implemented for VoIP to reduce jitter, delay, congestion and improve bandwidth utilization? Different Mechanisms have been developed in the past for QoS in communication networks and the fundamental objective is to avoid the congestion. QoS does not mean creating additional capacity but to prioritize certain type of traffic and to allocate the bandwidth in congested circumstances to maintain the efficiency of network . Advantages of VoIP are to reduce cost, efficient utilization of network resources, merging of voice and data traffic into a single network. These advantages cannot be achieved without QoS, understanding of legacy telephony system and knowledge of new VoIP technology. To achieve and study the effects of QoS mechanisms on VoIP and the background traffic effect on QoS, medium enterprise level of network will be designed by using OPNET simulation software.
Related Research and Relevant Technology
Voice is sensitive to delay and delay over 150ms mean loss of quality, however voice can tolerate loss while data traffic is sensitive to loss not to delay . These factors raise the issue of quality of service. Without QoS implementation, it will be the waste of network resources, equipment and company money. Research on QoS is one of the hot areas being carried out specially. VoIP through VPN (Virtual Private Network), security and its availability is another area gaining popularity and being researched 
Cisco has included QoS as a module for CCNP (Cisco Certified Network Professional) curriculum replacing old module of network troubleshooting. Well known QoS mechanisms are WFQ, FIFO, and CQ. Some of the call control models for VoIP are H.323, SIP (Session Initiation Protocol), MGCP (Media Gateway Control Protocol), Skinny (Cisco Unified Call Manager. Most commonly used are H.323 and SIP. H.323 is an old and quite complex protocol recommended by ITU-T while SIP is new and protocol of choice by most of the vendors and is recommended by IETF (Internet Engineering Task Force). Due to its simple and text based architecture SIP will be used as a main protocol for VoIP, home appliances, motor industry and many more .
- How does QoS mechanisms will influence the VoIP traffic behavior of medium enterprise level network?
- What will be the impact of background traffic on QoS?
Studying, understanding and designing of computer networks are becoming difficult . Development of communication protocols and increasing scale of networks are making it complex to install real world expensive networks. To overcome this problem network simulation tools such as OPNET, ns-2, CNET are being used. Main advantage of these tools is to model the real world system to predict the functionality, growth, problems without practical implementation of the actual equipment and making it a cost effective solution.
Medium enterprise level of network will be used to gather and analyse data based on bandwidth utilization, throughput and end-to-end delay in OPNET simulation software.
Medium Enterprise Level of Network.
Medium enterprise scale network will be designed with 10 routers, 200 workstations and 200 VoIP phones spread out on different subnets. Each subnet will consist of 20 workstations and 20 VoIP phones. Four application servers representing background traffic will be used, which are Database, HTTP, FTP and Email servers. Each subnet will be connected to a router via 24 port fast Ethernet switch. OSPF (Open Shortest Path First) will be the main routing protocol used across the network. For connectivity between workstations and switch, a link of 10BaseT will be used while switch, routers and application servers will be connected by 100BaseT link., to have more bandwidth available for the main network resources. Nine scenarios will be simulated on basis of different type of background traffic and different QoS mechanisms in the network i-e WFQ, FIFO and CQ. Results will be gathered in the numerical and graphical form for comparison and analysis purpose. Resource Requirements
Resource requirements are divided into two parts.
- Literature Review.
- Evaluation of dissertation.
Following methods will be used for literature review.
- Library Resources.
- Access to Internet.
- Access to Athens Journals.
- Access to IEEE Journals.
- Access to the university laboratory.
- OPNET simulation software to develop the composite network model.
- Initial meeting with the supervisor to discuss project schedule (1 day)
- Arrange schedule for the future meetings (1 day)
- Confirmation of necessary hardware and software (5 day)
- Carryout preliminary literature search (7 days)
- General OPNET review and availability (3 days)
- General design layout of project report (2 days)
- Two meetings with the supervisor to discuss the project in detail one in the start and one at the end of month (2 days)
- Detail literature search to collect as many resources as possible related to and helpful for the project (7 days)
- Review of material collected so far to be used in dissertation (4 days)
- Writing of dissertation report will continue (6 days)
- Creation and simulation of different topology layout in OPNET (4 days)
- Analysis of the results achieved from experiment (4 days)
- Meeting with the supervisor to discuss progress of the project so far and obtaining future directions (1 day)
- Project related material search and review (5 days)
- Design of network topology layout in OPNET and making necessary corrections (5 days)
- Troubleshooting OPNET designs and simulating network topologies (4 days)
- Analysis of the results achieved from experiment and finalizing them (3 days)
- Dissertation writing will continue (6 days)
- Conclusion of the dissertation along with appendixes (4 days)
- Review of the whole dissertation (5 days)
- Making final corrections necessary (4 days)
- Final meeting with the supervisor to review and evaluate the whole project (1 day).
- Creation of the presentation which covers everything (3 days)
- Submission of dissertation.
Evaluation of Dissertation
Following methods will be used for evaluation of dissertation.
This project is completely laboratory based and will be developed by using OPNET simulation software. It does not violate 1998 Data Protection Act. Project neither interfere potentially sensitive subjects nor human emotions.
Main risk factor is developing the composite network model by using OPNET simulation software accurately. OPNET does support the Project requirements. OPNET user tutorial guide, OPNET website, OPNET technical support and Internet will be used accordingly to understand and use simulation software accurately to minimize the risk factor and to get the optimum result.
- Voice over IP, (April 06, 2008). http://en.wikipedia.org/wiki/Voice_over_IP
- S. Sanghan. M. Hasan. "Intelligent P2P VoIP through Extension of Existing Protocols," School of Technology, Shinawatra University, Thailand, February 2007. P. 1597- 1601.
- D. Zhang and D. Ionescu. "QoS Performance Analysis in Deployment of DiffServ-aware MPLS Traffic Engineering" In Eighth ACIS International Conference on Software Engineering, Artificial Intelligence, Networking, and Parallel/Distributed Computing, Ottawa Canada, IEEE 2007. P.963-967.
- D. Orincsay. B. Jozsa. L. Tamasi. "Cost-optimal VPN based VOIP Network Design," Budapest University of Technology and Economics, Dept of Telecommunication and Media Informatics, 2002. P. 117-122.
- K. Wallace. "Cisco Voice over IP," Cisco Press, 2007. P. 283-284. N. AI-HoIou. "USING COMPUTER NETWORK SIMULATION TOOLS AS SUPPLEMENTS TO COMPUTER NETWORK CURRICULUM," 30th ASEE/IEEE Frontiers in Education Conference, Oct 2000. P.13-16.
- B. Morgan. N. Lovering. "CCNP ISCW Official Exam Certification Guide," Cisco Press, July 2007. P. 24-25.
- Ranjbar. "CCNP ONT Official Exam Certification Guide," Cisco Press, August 2007. P. 62-75.