Printer Friendly

Deciding Whether Your Local-Area Network Should Be PBX-Based or CPU-Controlled?

Many informatio systems managers are currently facing extremely difficult decisions relative to the development of a local-area network plan.

Both CPU and PBX vendors (and some that are in between) are courting users with wild-eyed promises of totally unlimited network functionality and open connectivity. The major vendors promise support for almost every protocol known to man and open access for a myriad of terminals, word processors, CPUs and other devices.

Users, however, must review all of the promises, evaluate the needs of their organizations, and make the choice of whether to use a computer-based broadband or PBX-based local-area network (LAN).

Since most users currently considering LANs have a substantial investment in PBXs and CPUs, the question becomes one of not only which vehicle can best act as the LAN controller, but also how the current investment in equipment can best be protected while migrating to an "open-access" environment. Evaluate Your Network Demands

To avoid the financial and operational problems of abandoning the current equipment environment and starting from scratch, users carefully evaluate their current and future network demands to develop a plan that will best serve their needs.

If such an evaluation reveals a heavy voice orientation in the network and/or the presence of a multitude of dispersed terminals needing access to multiple CPUs and each other, a PBX-based LAN may be the appropriate choice. If bursty, high-speed CPU-to-CPU traffic and/or the presence of few or clustered terminals exist, a CPU-controlled LAN may be the best option.

In our studies, the best method to adequately evaluate this question has been to construct multiple network models based on "what-if" considerations. Since the major differences between the two LAN schemes are operational speed and capacity, model development is imperative to adequately assess the particular situation.

Because much of the needed data is not readily available, it must be manually assembled or estimated. Although this tedious, time-consuming effort is an inexact science at best, it will pay dividends if used properly. Develop an Evaluative Model

The following case, involving a major user that's our client, will demonstrate the importance of developing an evaluate model.

This large user, like many others, has been struggling with the question of whether to utilize PBX systems as the control for a local-area network or to begin implementation of a computer-controlled LAN. This user currently has four major analog PBX systems serving approximately 7,000 voice terminals, and five major CPUs serving more than 1,000 inquiry/response terminals; all of this equipment is located within a four-square-mile campus configuration (see diagram). Establish Equipment Standards

Terminals presently are unable to access multiple CPUs, since they are either connected via channel attachment or multipoint private lines. Also located within this campus area is a mixture of remote job-entry terminals, personal computers and a wide range of word processors.

Standards have not been established for the acquisition of word processors and personal computers. This has resulted in a wide range of protocols and operational methods, making it impossible to share data or transfer documents. The situation is further complicated because the campus PBXs also act as switching points for a larger network. Formulate a Transitional Plan

To formulate a proper transitional plan for this user to migrate to a true "open-network" architecture, the volume and speed of each type of traffic were the most important factors to be considered. While long-term (more than 5 years) projections indicated that data traffic and CPU-to-CPU traffic would eventually exceed the volume of voice traffic, this is not the case today. Network traffic is currently 77 percent voice and only 23 percent data, with very little CPU-to-CPU applications in that distribution.

Based on the traffic model, we began the development of a possible implementation plan utilizing the PBX as the LAN controller. (A relevant assumption here is that PBX vendors will continue to improve the overall capacity and speeds of their networking systems. We feel this assumption to be well founded since data speeds of PBXs were limited to 9600 b/s just a few years ago and are now exceeding 56 kb/s, with even greater promises for the near future.) Divide the Plan into Phases

Once the decision that the network would utilize the PBX as its base was made, it then became necessary to develop a two-phase transitional plan for conversion.

The first phase would involve the conversion of the four analog PBXs to digital switches capable of handling the required data porting in addition to the required voice connections. At that time, it would also be necessary to redesign the intracampus network in order ot accommodate the traffic generated by the additional switched-data applications.

After carefully analyzing traffic and switching requirements, it was determined that three fourth-generation PBX switches (available from any one of at least four well-known vendors) would be adequate to handle the network requirements. These systems would be placed in Buildings 1, 2 and 3, and the analog switch in Building 6 would be replaced by an intelligent node from the switch in Building 2.

Off-premise extension (OPXs) in Buildings 4 and 5 would be replaced by intelligent nodes operating from the switches in Buildings 3 and 1, respectively. Fiber-optic cables would be used to connect the intelligent nodes, except in the case of Building 6, where three 1.544-Mb/s T1 circuits available from the local telephone operating company would be used to control costs.

The remaining local voice-grade trunks would also be replaced by T1-type circuits due to the expense of crossing public thoroughfares and the projected long-term transmission volumes. Eventually, these circuits would be replaced by fiber-optic cable, when the replacement cost justified it. Multiplexers Would Be Used

To terminate the T1 circuits, digital transmission multiplexers similar to the new Codex 6240 would be used. This type multiplexer offers voice and data integration over a single T1 circuit, and can be programmed to reconfigure transmission bandwidths dynamically based on network traffic parameters. They also provide multiple data transmission rates varying from 2400 b/s to 772 kb/s for data and rates of 16, 24, 32 and 64 kb/s for voice.

The next phase of the plan would call for the conversion of some 700 synchronous terminals to less-expensive asynchronous terminals. The PBX system would provide the protocol and speed conversion, thus eliminating the need for major CPU software conversions while improving data access and allowing word processors and personal computers to be connected to the network in a systematic, efficient manner. Savings Would Be Substantial

Estimated savings from the ability to utilize more cost-effective terminals and provide open connectivity to all the devices in the network is $650,000 over the next five years. In addition to these savings, the transition to T1 circuits would provide approximately $500,000 in savings over the same period.

Although this particular user scenario has certain unique aspects, it demonstrates the combined problem/opportunity facing many systems managers today, and a way to address it. While this process is not the one "perfect" approach, it does provide a systematic method for organizations to evaluate local-area networking options. Follow Recommended Guidelines

In conclusion to the process outlined above, we also recommend the following guidelines when considering an integrated network plan:

* Develop a combined communications department charged with the responsibility for all aspects of corporate communications.

* Carefully evaluate the direction of the company or entity relative to current and future network usage.

* Do not react to "state-of-the-art" prompts from vendors; the real art is to use current resources effectively.

* If you don't have network design expertise, hire a professional or retain a qualified consultant.
COPYRIGHT 1984 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1984 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Mecke, D.; Hill, J.
Publication:Communications News
Date:Oct 1, 1984
Words:1268
Previous Article:Xerox Finds Program Is TOPs for Private Network Savings.
Next Article:Environmental-Protection Pointers for Quality Teleconferencing-Room Design.
Topics:


Related Articles
Digital PBXs Playing Key Role in Today's Info Management, High-Speed Data Needs.
Pointers for Selecting Local Data Networks.
State Government Interfaces Call-Accounting Microcomputer to Mainframe Billing System.
Citibank wires intelligence into Hong Kong center.
Law firm reduces costs with networked fax.
Guidelines for planning wireless communications.
Prepared in paradise.
Prepared in paradise
CALISTA'S DPCP CONNECTS CISCO'S IP-PBX TO DIGITAL PHONES.
VoIP solutions--Get ahead of the game.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters