Roy Gedeborg

CSC 363

Mid-Term Exam

 

Report on Municipal Building Network

 

This report is split into four parts. The first item presented is a Proposal Summary Table that outlines the basic techniques and statistics which apply to the three proposals being presented. This summary does not attempt to outline advantages and disadvantages of each of the proposals. The following Technology Overview will cover the issues which separate the proposals. After the Overview, a detailed Analysis will describe the problems with each proposal and attempt to give options which might solve these problems. The final element of this report will be a Final Recommendation, which selects a plan that will best meet the needs of the new network.

 

Proposal Summary

 

Company making Proposal

Dune Networks

LHASA Technologies

FastCom

Type of Technology

FDDI

ATM

Fast Ethernet

Proposed Speed

100 Mbps

155 Mbps

100 Mbps

Cable Plant

Fiber and Cat-3

Fiber and Thick-net Coax

Cat-5

Internet Service

Direct Connection with Router/Concentrator & Gateway

Direct Connection with Router/Concentrator & Gateway

Direct Connection with Router/Concentrator & Gateway

Security Provided in Jail

Fiber

Fiber

Cat-5 cable

Proposed Routing Equipment

1 Concentrator

2 Concentrators

2 Switch/Active Hubs

Support Provided

Little

Little

3 Weeks of Training

 

Technology Overview

 

The fundamental differences in proposals lies in the networking medium. FDDI offers fiber security and competitive speeds over UTP. ATM offers the highest overall speed and fiber security as well as the ability to allocate bandwidth for streaming video teleconferences. Fast Ethernet brings the easiest network to maintain and troubleshoot and it also offers competitive speeds over the network. The differences in proposals also include technical support, with FastCom offering the best technical support, three weeks of side-by-side training. Security is also an issue, with excellent support provided by Dune and LHASA.

The three proposals are nearly identical in the Internet solution provided, a solid solution of a direct connection with a gateway and a router/concentrator a this connection. All proposals also provide weekend solutions to avoid inconveniences in installation. Finally all proposals provide a similar amount of connections, including two jacks to the fourth floor. However, the type of connections provided are far different in the service which they provide.

 

Proposal Analysis

 

The first network proposal, submitted by Dune Networks, uses FDDI technology to transmit data at 100 Mbps, which should be fast enough to support all of the situations needed in the building, including video teleconferencing. The network will run over multimode fiber on the bottom two floors and over Cat-3 cable on the upper three floors. This is the first, glaring error in Dune’s proposal. While Cat-3 cable is supported in a FDDI environment by 100 base T4 specifications, it is a dated cable plant and would provide little upward compatibility in future years. Cat-5 cable can be strung for little or no cost difference than Cat-5 cable. The added quality at no added cost seems to make it clear that Cat-5 cabling is the best choice in this decision which cannot be easily changed in the future.

The multimode fiber to be used with the network provides excellent security in a jail environment. This multimode fiber will be used to wire a dual-ring network for the bottom two floors and a concentrator will provide routing functions for the network. For the upper offices on the second and third floors, the stations will be single attached stations. No concentrator is provided for these stations. I am unsure on whether a single attached concentrator is required, but managing these single attached stations might be difficult if no concentrator exists. Therefore, it is highly recommended that some sort of single attached concentrator be added for the second and third floors in order to more easily manage the network in these areas.

The network operating system, Novell 4.12 has the capacity, capability and support needed in a network operating system. The Internet specifications for Dune, as well as all other proposals is broad enough, with a gateway and router/bridge specified to connect to a direct line from an Internet provider. Additionally, Dune agrees to perform all work over weekends, just like the proposals from the other two companies. Finally, Dune hasn’t offered much in the way of technical support once the network is up and running.

The network proposal submitted by LHASA Technologies uses ATM technology as its media. ATM will be run using multimode fiber over a dual attached ring at speeds up to 155 Mbps. This is the most impressive speed submitted by any of the proposals, and ATM is by far the best media for video teleconferencing because it allows allocation of bandwidth. In addition, the bottom two floors will have backbone speeds brought to every device!

The attractiveness of this technology is offset by a very poor decision in the cable plant to the second and third floors of the building. Thick-net coaxial cable is a slow, primitive Ethernet technology which won’t transmit at speeds much over 10 Mbps. The most worrisome thing about LHASA’s proposal is the line which states, "Coax also provides an excellent medium for video delivery." It is true that video conferencing can be done at 2-6 Mbps with high compression techniques, but that type of bandwidth would bring the network to its knees on the upper floors, not allowing any other activity to occur while video conferencing took place. In order to provide good service, as well as options to expand (for such things as future HDTV transmissions), it would be much better to string Cat-5 Unshielded Twisted Pair. Not only will this provide for much, much higher bandwidth transmission, but it would also be far less expensive ($80-$100/500 ft while Thick-net costs at least $300/500 ft.). Of course, 2-3 switch/active hubs and a bridge/router will need to be purchased in order to run the network over Cat-5, but the difference in cable costs will make the two options comparable in price. In addition, the bridge/router is the same device which would be needed (and that LHASA didn’t provide for) to run the Thick-net network at far slower speeds.

The network operating system, Windows NT 5.0, should provide a good operating system, although I would question its maturity as NT 4.0 has needed several service packs, patches, and fixes in order to cover some of its flaws. Still, NT does offer the best operating system features, generally improving on Novell’s service and support in my opinion. The two concentrators should provide good service to the dual-attached stations on the lower two floors. Multimode fiber provides great security in the jail environment too. The Internet solution and the willingness to work on weekends puts LHASA’s proposal on the same level as the other proposals. Finally, LHASA has offered no real technical support, but with such poor choices of media, it is questionable whether or not any technical support from this company would be worth the trouble.

The third proposal submitted to the County Building by FastCom, uses Fast Ethernet over Cat-5 cable as its media and technology. Fast Ethernet offers 100 Mbps speed and a star topology which allows for the easiest station switching out of all of the proposals. In fact, FastCom’s proposal seems to be the most sound out of the group, showing no huge errors in concept and design. In addition, Ethernet is my personal favorite choice for network design because it utilizes as much bandwidth as possible rather than providing dedicated bandwidth like ATM does. However, this approach will not work as well for video teleconferencing or other streaming data applications being performed over the network. In addition, Cat-5 cable can be tapped without detection; a concern in the jail environment, but not a large concern if proper monitoring is done of both prisoners and visitors.

In my estimation, one minor error was made by FastCom’s proposal. Two pairs of Cat-5 wire are specified to be brought to each desktop. I think that the full four pairs should be brought to each jack because the cost is relatively minimal and it will be much easier to upgrade in the future to faster technologies without having to string extra cable through the walls. The network operating system proposed by FastCom, Novell Netware 4.12, is an industry standard with a lot of support and capabilities, making it a solid choice. The Internet proposal and ability to work on weekends are the same as FastCom’s competitors. Finally, FastCom offers terrific networking support, including three weeks of side-by-side training. The sound reasoning behind the proposal and the excellent features provided by FastCom make this proposal an attractive one.

 

Final Recommendation

 

After having dissected the options available in the previous analysis, I think that the best two proposals are offered by FastCom and LHASA. The superior technology of ATM makes LHASA’s proposal intriguing, but the technical knowledge demonstrated by the proposing group is not at the level that would be desired. FastCom provides a competitive proposal and the technical knowledge and support needed for new networks, but their Fast Ethernet technology does not have the speed or design which would allow high quality, multiple video teleconferences for many years.

Therefore, it is my recommendation that we check into FastCom’s ability to service our proposal with ATM technology. Their professional knowledge and technical support would be a valuable commodity for installing a new ATM network. However, if FastCom does not feel comfortable in implementing ATM technology, I would recommend hiring LHASA and looking for some sort of outside consultant to overlook installation and provide technical support for a couple of weeks after the network is finished. In addition, I would also echo FastCom’s sentiments that at least one full-time network administrator be hired to take care of the new investment being made.

If LHASA is given the project, I would specify Cat-5 cabling to be brought to the upstairs floors rather than Thick-net Coax. In addition, I would also check into the possibility of providing for a direct fiber connection to the video-teleconferencing area on the second floor for the Health Department Conferencing Room. Even if it is determined that such a connection is not necessary for teleconferences today, I would specify that fiber be brought to all but the very top floor in order to expedite future upgrade cabling jobs to this area.

With this solution, I feel that teleconferencing, our most demanding need, can be adequately met. In addition, excellent security, the best speed, and strong control of network resources would also be a part of the new network. Internet capability and technical support would be two secondary considerations which are more than adequately met by this recommendation. Finally, if pinching pennies becomes necessary, the network should be implemented partially on the upper two floors. Full networking capabilities can be added later, but a good strong network should not be compromised just in order to reach every last station in the building.

 

Network Diagram

 

3rd Floor offices and Machines 1st Floor Offices and Machines

Active Hubs Concentrators =>

2nd Floor offices and Machines Basement Machines

 

Internet Gateway