Analysis of a trunked radio system utilization using real billing traces:

Presentation slides and final report (PDF files).

The goal of this project is to explore and analyze the current system utilization of a real, public safety communication network, operated by E-Comm. The basic network consists of a central site and a number of cells, each cell having different capacity expressed in terms of number of available radio channels. Since the network is basically circuit-switched, the system utilization can be viewed as a time and space distribution of the number of concurrent calls. From the available traffic traces it is possible to create a model in OPNET that will run using these traces. After creating the model and running the simulation, we expect to draw conclusions about overall utilization of system resources and to locate existing or possible bottlenecks. This is particularly important if the amount of traffic increases, which is a reasonable asumption. Using these results, it is possible to propose ways of solving existing and future network congestion problems.

References:
[1] L. A. Andriantiatsaholiniaina and L. Trajkovic: "Analysis of User Behavior from Billing Records of a CDPD Wireless Network", Workshop on Wireless Local Networks (WLN) 2002, Tampa, FL, Nov. 2002, pp. 781-790 http://www.ensc.sfu.ca/~ljilja/papers/WLN02_Andriantiatsaholiniaina.pdf
[2] L. Irwing "Land Mobile Spectrum Planning Options (report)", National Telecommunications and Information Administration (NTIA), 1995, (APPENDIX: Sharing Trunked Public Safety Radio Systems Among Federal, State, and Local Organizations) http://www.ntia.doc.gov/osmhome/reports/slye_rpt/appendix.html
[3] S. Panossian and D. Medhi: "Towards Providing Enhanced 911 Emergency Service in IP Telephony", Department of Computer Networking, University of Missouri, Kansas City, 1999 http://www.cstp.umkc.edu/public/papers/dmedhi/pm_tr_22r_99.pdf
[4] EDACS Trunking Information: http://www.trunkedradio.net/modules.php?name=Content&pa=showpage&pid=2
[5] OpenSky Network: http://www.opensky.com/network/index.asp

Implementation of a new Queueing Algorithm using Fuzzy Logic:

Presentation slides and final report (PDF files).

This project investigates a new Queueing Algorithm using Fuzzy Logic to implement congestion control and to utilize as fully the resources as possible.

In a typical network, many computers are connected to a central queue, as in the case of an ISP. If the sum of the traffic rate of the end computers exceed the processing capability of the queue, there must exists a method for which the queue shares its total processing capability with each computer according to a set of pre-defined critieria.

The results from this project will be compared to existing alogirthms in terms of throughput, delay/jitter, and fairness.

References:
[1] A fuzzy admission control scheme for high-speed networks Barolli, L.; Koyama, A.; Yamada, T.; Yokoyama, S.; Suganuma, T.; Shiratori, N.; Database and Expert Systems Applications, 2001. Proceedings. 12th International Workshop on , 2001 Page(s): 157 ?61
[2] Design of a fuzzy traffic controller for ATM networks Ray-Guang Cheng; Chung-Ju Chang; Networking, IEEE/ACM Transactions on , Volume: 4 Issue: 3 , Jun 1996 Page(s): 460 ?69
[3] Fuzzy admission control and scheduling of production systems Runtong Zhang; Phillis, Y.A.; Robotics and Automation, 1999. Proceedings. 1999 IEEE International Conference on Volume: 2 , 1999 Page(s): 1508 -1513 vol.2
[4] A fuzzy-based traffic controller for high-speed ATM networks using realistic traffic models Kasiolas, A.; Makrakis, D.; Multimedia Computing and Systems, 1999. IEEE International Conference on , Volume: 2 , Jul 1999 Page(s): 389 -394 vol.2
[5] Admission control and scheduling in simple series parallel networks using fuzzy logic Runtong Zhang; Phillis, Y.A.; Fuzzy Systems, IEEE Transactions on , Volume: 9 Issue: 2 , Apr 2001 Page(s): 307 -314

Analysis of voice communications in IEEE 802.11 wireless networks:

Presentation slides and final report (PDF files).

This paper investigates Quality of Service (QoS) of Voice over IEEE 802.11 wireless networks. The OPNET software package is used to simulate and determine the number of mobile users/access points that meet acceptable QoS parameters. Depending on the packet size, we will determine the maximum range of background voice users that meet an acceptable QoS. QoS for voice over IEEE 802.11 can be determined directly from packet loss and packet timing jitter. Typical acceptable values of packet loss and packet timing jitter are 1% loss and 100-200 ms delay respectively.

References:
[1] Brian P. Crow, Indra Widjaja, Jeong Geun Kim, Prescott T. Sakai, IEEE 802.11 Wireless Local Area Networks, IEEE Communications Magazine, pp. 116-126, September 1997.
[2] Feigin, J., Pahlavan, K., Ylianttila, M., Hardware-Fitted Modeling and Simulation of VoIP over a Wireless LAN, VTC 2000, Volume 3, pp. 1431-1438, Fall 2000.
[3] Visser, M.A., El Zarki, M., Voice and Data Transmission over an 802.11 Wireless Network, PIMRC'95, Volume 2, pp. 648-652, September 1995.
[4] Sobrinho, J.L., Krishnakumar, A.S., Distributed Multiple Access Procedures to Provide Voice Communications over IEEE 802.11 Wireless Networks, GLOBECOM'96, Volume 3, pp. 1689-1694, November 1996.
[5] Prasad, A.R., Performance Comparison of Voice over IEEE 802.11 Schemes, VTC 1999, Volume 5, pp. 2636-2640, Fall 1999.

Gnutella Connection Dynamics:

Presentation slides and final report (PDF files).

Gnutella is a peer-to-peer protocol designed for distributed file sharing. Unlike other file sharing networks, such as the now defunct Napster, Gnutella is a decentralized model in which every node, termed a "servent," is both a server and a client. Due to its distributed nature, and the rate that servents tend to join and leave the network, the topology of the network is very dynamic.

For a new node to join the network, it must first contact a stable central server (such as router.limewire.com) to find the IP addresses of some nodes currently in the network. Then, the servent connects to these nodes in the hope of finding one that is accepting new connections. When such an opportunity for a new connection is found, the servent links into the network and may query and share files.

This method of connecting to the network is rather slow in practice. Each time a new servent wants to join the network, it must find a node that is accepting new connections before it can participate in the network. This process tends to take on the order of minutes. Moreover, the usefulness of the network is directly related to stability of the connections around a node, and the number of nodes that a user's queries can reach. Each of these properties is based on the stability and diversity of the connections a node makes with the rest of the network.

Therefore, we intend to investigate the dynamic formation of network connections in the Gnutella network. We will begin by investigating the details of how new servents currently connect to the network. We will investigate why it takes on the order of a minute or more to establish a connection, and look at ways to reduce this time. A second, and related objective is to investigate how a node currently connected to the network could create a connection to a node in a distant part of the network. This would allow such a node to search a much greater number of other nodes for any given query, and thus allow a user to find the more rare resources on the network.

We will implement the current connection creation mechanism, as well as our own improved approach using the NS-2 network simulator. We will compare both the time required to create a connection, and the diversity of the connections created. We expect our simulations to show that our new approach will yield shorter connection times and a better distribution of network connections.

References:
1. Bildson, G. "PROPOSAL: Handshaking Protocol", The Gnutella Developer Forum (GDF), 2001. http://groups.yahoo.com/group/the_gdf/message/2010
2. Ripeanu, M. "Peer-to-Peer Architecture Case Study: Gnutella Network", Proceedings of the First International Converence on Peer-to-Peer Computing (P2P'01), 2002.
http://people.cs.uchicago.edu/~matei/PAPERS/gnutella-rc.pdf
3. Rohrs, C., "Query Routing for the Gnutella Network", 2002, May 16, www.limewire.com/developer/query_routing/keyword%20routing.htm
4. Rohrs, C., Singla, A. "Ultrapeers: Another Step Towards Gnutella Scalability" 2002, Nov. 26, www.limewire.com/developer/Ultrapeers.html
5. Rohrs, C. "PingPong", www.limewire.com/index.jsp/pingpong

Introduction of TCP packet queue in WAPs to improve TCP preformance:

Presentation slides and final report (PDF files).

The project investigates queuing of TCP packets at wireless access points (WAPs) destined for wireless nodes. Currently, if a WAP looses access to a mobile node, the TCP packets get dropped until communications is reestablished with the node a t which point communications is resumed. We model the introduction of different queue sizes inside the WAP which will acce pt TCP packets if the signal to a node is lost. If communications is reestablished before the queue runs out then the WAP can transmit the queued packets to the node a lot faster than if the node re-requested them from the source. Additionally, it would make sense to have an additional functionality put into the WAPS. The functionality would allow WA P-A to forward TCP packets to a WAP-B if WAP-A looses communications with a mobile node, which eventually enters the commu nications range of WAP-B. As soon as the network knows that the Node is in contact with WAP(B) [taken care of the protocol ], the QUEUE of WAP-A gets forwarded to WAP-B and then to the Node.

References:
- Efficient TCP over Networks with Wireless Links by Elan Amir, Hari Balakrishnan, Srinivasan Seshan, Randy H. Katz; Computer Science Division, University of California at Berkeley
- TCP over Wireless by Ramesh Neelamani and Amit Saxena; Department of Electrical and Computer Engineering, Rice University, Houston Texas
- Efficient Information Access for Wireless Computers by Stuart Brian Wachsberg; Thesis from University of Waterloo
- Improving Reliable Transport and Handoff Performance in Cellular Wireless Networks by Hari Balakrishnan, Srinivasan Seshan, and Randy H. Katz; Computer Science Division, Department of Electrical Engineering and Computer Science, University of California at Berkeley, Berkeley
- A Survey of TCP Performance on Wireless Links by Brooke Shrader; Royal Institute of Technology (KTH), Stockholm, Sweden
- Brooke Shrader, "A Survey of TCP Performance on Wireless Links" March 11, 2002
- M. N. Mehta, N. H. Vaidya, "Delayed Duplicate-Acknowledgements: A Proposal to Improve Performance of TCP on Wireless Links", 1998
- S. Dawkins, G. Montenegro, M. Kojo, V. Magret, N. Vaidya, "End-to-end Performance Implications of Links with Errors", IETF RFC 3155, August 2001

Comparing End System Multicast and Traditional Streaming Data Distribution Using OPNET:

Presentation slides and final report (PDF files).

End System Multicast (ESM) allows for live audio and video content to be delivered over the Internet at a low cost to the broadcaster. It achieves this by harnessing the idle network resources of its viewers to help distribute the broadcast. The viewers automatically arrange themselves into a broadcast distribution tree. The tree continuously adapts to network dynamics, attempting to obtain the highest bandwidth and lowest latency.” In this project we will simulate both ESM and traditional model of multicast streaming in which each client connects directly to the server and compare the results.

References:

[1] Y. Chu, S. Rao, S. Seshan and H. Zhang, ``A Case for End System Multicast&rdquo,'' IEEE Journal on Selected Areas in Communication (JSAC), 2002: http://www-2.cs.cmu.edu/~narada/JSAC/jsac.ps, (Mar. 2003).

[2] Y. Chu, S. Rao, S. Seshan and H. Zhang, ``Enabling Conferencing Applications on the Internet using an Overlay Multicast Architecture,'' Proceedings of ACM SIGCOMM, 2001: http://www-2.cs.cmu.edu/~narada/Sigcomm2001/paper.ps, (Mar. 2003).

[3] Y. Chawathe, ``Scattercast: An Architecture for Internet Broadcast Distribution as an Infrastructure Service,'' PhD Thesis, University of California at Berkeley, 2000: http://berkeley.chawathe.com/thesis/thesis-single.ps.gz, (Mar. 2003).

[4] J. Jannotti, D. Gifford and K. Johnson, ``Overcast: Reliable Multicasting with an Overlay Network,'' Cisco Systems, 2000: http://www.jannotti.com/papers/overcast-osdi00.ps, (Mar. 2003).

Handover in GPRS networks:

Presentation slides and final report (PDF files).

The GPRS standard offers data transfer rates that render possible a number of new and interesting applications for mobile terminals. This project will elaborate on the ability of a GPRS network to meet the requirements of an agreed QoS profile while carrying out a handover. A GPRS network, capable of performing handover, will be implemented and simulated in the network modeling tool "Opnet".

References:
[1] General Packet Radio Service (GPRS), Service description, 3GPP TS 03.60 version 7.9.0 Release 1998
[2] General Packet Radio Service (GPRS), Overall description of the GPRS radio interface, 3GPP TS 03.64 version 8.10.0 Release 1999
[3] General Packet Radio Service (GPRS), Radio Link Control/Medium Access Control (RLC/MAC) protocol, 3GPP TS 44.060 version 4.9.0 Release 4
[5] General Packet Radio Service (GPRS), External network assisted cell change (NACC), 3GPP TR 44.901 version 5.1.0 Release 5

Simulation and Analysis of Wrap Protection in Resilient Packet Rings (RPR):

Presentation slides and final report (PDF files).

Since the Internet has become a fundamental part of society, network resiliency and availability has become the most important priority for service providers, as disruptions to a network will result in loss of business for both its users and providers.

Resilient Packet Ring (RPR) is a technology being developed as a scalable and resilient link layer solution for metropolitan area networks (MANs). RPR is currently being drafted as IEEE standard 802.17. RPR implements a dual counter rotating ring topology with global statistically-multiplexed allocation of bandwidth that has the high fault tolerance of circuit-switched networks but is optimized for packet transmission. The goal of RPR is to be able to restore network operations in less than 50 ms from the occurrence of a link failure. To achieve this, RPR uses a number of protection mechanisms. One such mechanism is wrap protection, in which the nodes on either side of a failed link automatically wraps the traffic from one ring to the other.

As part of the project, OPNET will be used to model an RPR network, and to simulate the wrap protection mechanism. Measurements will be taken of the packet count and end-to-end delay over a short period of time to verify whether the requirement of a less than 50 ms failover is achievable.

References:
1) Tsiang, D, G. Suwala, "The Cisco SRP MAC Layer Protocol", RFC 2892, The Internet Engineering Task Force, August, 2000, http://www.ietf.org/rfc/rfc2892.txt?number=2892
2) "Spatial Reuse Protocol Technology", Cisco 7200 Series Router Product White Paper, Cisco Systems, 2003,
3) "A Summary and Overview of the IEEE 802.19 Resilient Packet Ring Standard", Resilient Packet Ring Alliance, 2003, http://www.rpralliance.org/articles/overview_of_draft_2.pdf
4) "An Introduction to Resilient Packet Ring Technology", Resilient Packet Ring Alliance, 2003, http://www.rpralliance.org/articles/ACF16.pdf
5) "Outline of the IEEE 802.17 RPR Draft Standard", Resilient Packet Ring Alliance, 2003, http://www.rpralliance.org/articles/ACF18.pdf

Comparison of Bluetooth Scatternet Formation Algorithms:

Presentation slides and final report (PDF files).

This project will focus on two primary aspects of the simulation of a Bluetooth network in ns-2. The first is the implementation of an improved and more efficient algorithm of Bluetooth scatternet formation than that found in the Blueware Bluetooth simulator for ns-2. The second is development of a module to implement a radio propagation model to account for signal interference and energy dissipation over the wireless interface, which has not been implemented in the Blueware models.

References:
[1] Kulwinder Atwal and Ron Akers, "Transmission of IP Packets over Bluetooth Networks", http://www.globecom.net/ietf/draft/draft-akers-atwal-btooth-01.html.
[2] A. Karnik and A. Kumar, "Performance analysis of the Bluetooth physical layer," IEEE Intl. Conf. on Wireless Personal Commun. (ICPWC'2000).
[3] C. Law and K.-Y. Siu, "A Bluetooth Scatternet Formation Algorithm", Proc. GLOBECOM '01,Vol.5, pp. 2864--2869, (2001).
[4] Godfrey Tan. "Self-organizing Bluetooth Scatternets", SM Thesis, Massachusetts Institute of Technology, January, 2002.
[5] Godfrey Tan, Allen Miu, John Guttag, and Hari Balakrishnan. "An Efficient Scatternet Formation Algorithm for Dynamic Environments", IASTED Communications and Computer Networks (CCN), Cambridge, MA, November, 2002.

Endpoint Admission Control implemented in OPNET:

Presentation slides and final report (PDF files).

In order to provide QoS to the Internet and avoid scalability problems, several recent papers proposed endpoint measurement-based admission control (MEAC) schemes.In these designs, before transmitting real data, endpoints probe the path by sending probing packets at the data rate it would like to consume. The connection will be established if the probe experiences a quality above a specified threshold. This mechanism does not relay on any additional procedure in internal network routers other than the capability of providing differentiated services. Analysis and simulation results mentioned in these papers suggest that a soft real-time service can be supported by EMAC. If this method is successful, it would represent a dramatic shift in the way QoS is managed in todayˇŻs Internet.

Our project is devoted to the study of the EMAC scheme. We will implement and simulate the basic EMAC scheme in OPNET. We will create the model of the endpoint, which can probe the network and make an admission decision. We will modify the existing model ¨Cpriority queue in OPNET so that the model can provide service to different traffic classes with a rate limiter. Corresponding packet formats will be defined to stand for probing, data and feedback packets. Several simulation scenarios based on different traffic source, traffic load and probing time will be explored.

References:

QoS at the diffserv network edge:

Presentation slides and final report (PDF files).

At present, the Internet lacks the QoS (Quality of Service) that is required to support delay sensitive traffic such as real-time voice. The original design and development of IPv4 (Internet Protocol version 4) did not consider that this "best-effort" packet delivery network might be one day used do accommodate real-time CBR traffic such as voice flows. Even today, the Internet does not provide QoS.

Backbone network providers should not and need not be burdened with issues of QoS and should be able to rely on their respective ISPs to carry out the required traffic engineering based on backbone/ISP SLAs (service level agreements). To achieve QoS in the Internet, traffic engineering is required at access points from the customer network to the ISP. Due to the administrative burden placed on network operators by intserv (integrated services such as RSVP), I will explore strategies which will place intserv functions within the responsibility of the customer network only. The ISP network will provide the required levels of QoS by the exclusive use of diffserv strategies. These strategies will include traffic policing and any necessary re-classification based on flow identification (IP address / port identification), packet size and other packet or flow characteristics.

In this project I will investigate queuing strategies involving the use of diffserv or DSCP (differentiated services code point) implementation. Specifically, there is a requirement to police, possibly re-classify and queue packet traffic at the ISP border/access router to meet QoS requirements.

References:
1] B. Braden, D. Clark, J. Crowcroft, B. Davie, S. Deering, D. Estrin and S. Floyd, "Recommendations on Queue Management and Congestion Avoidance in the Internet," IETF RFC-2309, http://www.zvon.org/tmRFC/RFC2309/Output/frontpage.html (current April 1998).
[2] Y. Bernet, P. Ford, R. Yavatkar and F. Baker, "A Framework for Integrated Services Operation over Diffserv Networks", IETF RFC-2998, http://www.zvon.org/tmRFC/RFC2998/Output/frontpage.html (current November 2000).
[3] Tomi Solala, "A Framework for Integrated Services over Diffserv Network", http://www.cs.helsinki.fi/u/kraatika/Courses/QoS00a/solala.pdf (currrent November 2000)
[4] K. Nichols, V. Jacobson, L. Zhang, "A Two-bit Differentiated Services Architecture for the Internet", IETF RFC-2638, http://zvon.org/tmRFC/RFC2638/Output/frontpage.html, http://irl.cs.ucla.edu/papers/twobit.pdf (current July 1999).
[5] Avaya Inc., "Avaya IP Voice Quality Network Requirements," http://www.ipintegration.co.uk/PDFs/IP Network Requirements.pdf (current June 2001).
[6] Cisco Press, Cisco IOS 12.0 Quality of Service, Cisco Press, Indianapolis, IN, 1999, pp. 87-98 (chpt. 11)
[7] Jonathan Davidson and Tina Fox, Deploying Cisco Voice over IP Solutions, Cisco Press, Indianapolis, IN, 2002, pp. 61-108 (chpt. 3)
[8] http://www7.homeunix.net/voip/10_3_labs(1-12)instructor.pdf

Adaptive Gossip-Based Ad Hoc Routing Algorithm:

Presentation slides and final report (PDF files).

A mobile ad hoc network (MANET) is a system of wireless mobile nodes that does not rely on any base station or fixed infrastructure. It poses difficult challenges on routing protocols due to multi-hop wireless connectivity, frequently changing network topology and the need for efficient dynamic routing protocols. Gossip-based technique recently employed in Ad Hoc routing algorithm has achieved significant improvement on both routing efficiency and reliability. We implemented an existing gossip-based Ad Hoc routing algorithm added in Ad Hoc On-Demand Distance Vector Routing (AODV+G), and found some problems of this gossip-based algorithm that can be solved to further improve the overall routing performance. A new algorithm is introduced, simulated and compared with AODV+G and results in significant performance improvement. The simulation model with MAC and physical layer models is used to study interlayer interactions and their performance implications.

References:
1. Rajendra V. Boppana and Satyadeva Konduru, An adaptive distance vector routing algorithm for mobile, ad hoc networks. Proceedings of the Twentieth Annual Joint Conference of the IEEE Computer and communications Societies, pp. 1753-1762, vol 3, 2001.
2. T. Camp and J. Boleng and B. Williams and L. Wilcox and W. Navidi, Performance comparison of two location based routing protocols for ad hoc networks, Proceedings of IEEE INFOCOM, pp.1678-1687, 2002.
3. author = L. Li and J. Halpern and Z. Haas, Gossip-based Ad Hoc Routing,L. Li, J. Halpern, Z. J. Haas, Gossip-based Ad Hoc Routing, unpublished. citesser.nj.nec.com/hass02gossipbased.html.
4. Jorjeta Jetcheva and David B. Johnson, Adaptive Demand-Driven Multicast Routing in Multi-Hop Wireless Ad Hoc Networks, Proceedings of the Second Symposium on Mobile Ad Hoc Networking and Computing, Oct. 2001.
5. Yih-Chun Hu and Adrian Perrig and David B. Johnson, Ariadne: A Secure On-Demand Routing Protocol for Ad~Hoc Networks, The 8th ACM International Comference on Mobil Computing and Networking, Sep. 2002. citeseer.nj.nec.com/bu02ariadne.html.

Simulation and Analysis of Content Delivery Network:

Presentation slides and final report (PDF files).

Traditional caching has limited effectiveness due to various reasons, and additional mechanisms are needed to deliver acceptable performance.

A Content delivery network(CDN) maintains multiple locations with copies of the same content, and uses information about the user and the content requested to "route" the user to the most appropriate site, thus offload work from original servers.

The project of our team is to implement a simple Content Delivery Network framework. We mainly interested in two techniques: re-direction and load-balancing. We may need to modify the DNS-related protocols and make them serve CN. Also if time permitted, we will experiment with some security technique in the Content Network.

We plan to use OPNET to perform the simulation.

References:
1. Rajendra V. Boppana and Satyadeva Konduru, An adaptive distance vector routing algorithm for mobile, ad hoc networks. Proceedings of the Twentieth Annual Joint Conference of the IEEE Computer and communications Societies, pp. 1753-1762, vol 3, 2001.
2. T. Camp and J. Boleng and B. Williams and L. Wilcox and W. Navidi, Performance comparison of two location based routing protocols for ad hoc networks, Proceedings of IEEE INFOCOM, pp.1678-1687, 2002.
3. author = L. Li and J. Halpern and Z. Haas, Gossip-based Ad Hoc Routing,L. Li, J. Halpern, Z. J. Haas, Gossip-based Ad Hoc Routing, unpublished. citesser.nj.nec.com/hass02gossipbased.html.
4. Jorjeta Jetcheva and David B. Johnson, Adaptive Demand-Driven Multicast Routing in Multi-Hop Wireless Ad Hoc Networks, Proceedings of the Second Symposium on Mobile Ad Hoc Networking and Computing, Oct. 2001.
5. Yih-Chun Hu and Adrian Perrig and David B. Johnson, Ariadne: A Secure On-Demand Routing Protocol for Ad~Hoc Networks, The 8th ACM International Comference on Mobil Computing and Networking, Sep. 2002. citeseer.nj.nec.com/bu02ariadne.html.

Efficiency of Packet Fragmentation in IP:

Presentation slides and final report (PDF files).

Packet fragmentation in IP networks occurs when source packets are split into smaller packets in order to fit the maximum size of some link along the route. When done "on-the-fly", this splitting can be quite computationally expensive for the router performing the fragmentation. Shannon, Moore, and Claffy in IEEE/ACM Transactions on Networking Dec 2002 ("Beyond Folklore: Observations on Fragmented Traffic") document that Microsoft Media Player is one of the leading causes of fragmented traffic on the public IPv4 backbones. We will study the impact this fragmentation has on the performance of the media stream in different network configurations (one "worst-case" scenario occurs when the first router has to fragment every packet being send by a server in heavy use) and compare it with end-point fragmentation as will be mandated in IPv6.

References:
Beyond folklore: observations on fragmented traffic Shannon, C.; Moore, D.; Claffy, K.C.; Networking, IEEE/ACM Transactions on, Volume: 10 Issue: 6 , Dec 2002 Page(s): 709-720
http://ieeexplore.ieee.org/iel5/90/25183/01134297.pdf?isNumber=25183&prod=IEEE+JRN&arnumber=1134297&arSt=709&ared=720&arAuthor=Shannon%2C+C.%3B++Moore%2C+D.%3B++Claffy%2C+K.C.%3B
RFC 791 - Internet Protocol, Darpa Internet Program Protocol Specification September 1981 http://www.faqs.org/rfcs/rfc791.html
Fragmentation Considered Harmful Kent, C. A. and Mogul J. C. Proc. SIGCOM '87 Vol. 17, Num. 5 October 1987 http://doi.acm.org/10.1145/205447.205456
RFC 2460 Internet Protocol, Version 6 (IPv6) Specification December 1998 http://www.faqs.org/rfcs/rfc2460.html
RFC 768 User Datagram Protocol August 1980 http://www.faqs.org/rfcs/rfc768.html

IP Classification and Metering:

Presentation slides and final report (PDF files).

High speed Internet Service Providers (ISPs) in the past offered shared bandwidth to thousands of subscribers. These ISPs provided a flat rate subscription fee to all subscribers regardless of bandwidth usage. As a result, a small percentage of these subscribers generated most of the bandwidth going through the ISPs backbone connection. Ideally, ISPs want to have a positive correlation between bandwidth usage and new customer subscription. In order to maximize the revenue earned from a given backbone connection, ISPs implemented a multi-tier subscription level where each tier will be guaranteed to a specific bandwidth at any one time.

To guarantee certain bandwidth to different customers, IP classification and metering algorithms will have to be implemented. Based on a customer's subscription level, IP packets received are processed (passed or dropped) depending on network usage. Network usage is determined by the bandwidth usage from customers with different subscription level. We will be simulating the traffic activity of a neighborhood where different customer premises have different subscription level. Through simulation, we will determine whether our implementation can provide customers with the guaranteed service that is associated to their subscription level.

References:
[1] Internet Protocol DARPA Internet Program Protocol Specification, RFC791, September 1981, http: //www.ietf.org/rfc/rfc0791.txt?number=791
[2] Amitava Dutta-Roy, An Overview of Cable Modem Technology and Market Perspective, IEEE Communic ations Magazine, pp 81-88, June 2001
[3] David Fellows, Doug Jones, DOCSIS Cable Modem Technology, IEEE Communications Magazine, pp 202 -209, March 2001
[4] Peter Komisarczuk, IP Access Service Provision for Broadband Customers, The Institution of Ele ctrical Engineers, 1999
[5] IP Quality of Service:An Overview, http://qos.ittc.ukans.edu/ipqos/ip_qos.htm

Performance Analysis of Megaco/H.248 Protocol over ATM and IP Network Using OPNET:

Presentation slides and final report (PDF files).

The convergence of voice and data networks is the driving force behind the media gateways and media gateway controllers. Such gateways allow the streaming of voice, video and data across a single network infrastructure. To initiate the communications between a media gateway and media gateway controller, Internet Engineering Task Force (IETF) and International Telecommunication Union (ITU) developed the Megaco/H.248 protocol together in Year 1998.

In our project, we will extend the Megaco/H.248 simulation, which was implemented by Riadul Mannan, Mahmood Riyadh and Shufang Wu in Spring 2002, to incorporate all the Megaco commands using OPNET. In addition, an analysis will be carried out to compare the performance of the Megaco/H.248 Protocol running on top of the ATM and IP network.

References:
[1] Tom Taylor, "Megaco/H.248: A New Standard for Media Gateway Control", IEEE Communications Magazine, pp. 124-132, October 2000.
[2] N. Greene, M. Ramalho, and B. Rosen, "Media Gateway Control Protocol Architecture and Requirements", RFC 2805, April 1999. http://www.ietf.org/rfc/rfc2805.txt, accessed in Febrary 2003
[3] F. Cuervo, N. Greene, A. Rayhan, C. Huitema, B. Rosen, and J. Segers, "Megaco Protocol Version 1.0", RFC 3015, November 2000. http://www.ietf.org/rfc/rfc3015.txt, accessed in Febrary 2003
[4] P. Blatherwick, R. Bell, and P. Holland, "Megaco IP Phone Media Gateway Application Profile", RFC 3054, January 2001. http://www.ietf.org/rfc/rfc3054.txt, accessed in Febrary 2003
[5] M. Brahmanapally, P. Viswanadham, and K. Gundamaraj, "Megaco/H.248 Call flow examples", October 2002. http://www.ietf.org/internet-drafts/draft-ietf-megaco-callflows-01.txt, accessed in Febrary 2003