Is there a MAN in your future?
We've all had our fun with integrated services digital networks. ISDN descriptions range from "I Still Don't Know" to "I Smell Dollars Now." A few of you who read this column have installed ISDN. I've talked to many of you about it. For most of you, the concept--and the service--has worked reasonably well. It fits into your wide area network (WAN) plans. And you've been able to work around the cost issues.
But now we face a new technology on the horizon: metropolitan area networks, or MANs. Although several definitions for a MAN have been floated, let's try our own: A MAN provides high-speed, switched network connections between high-speed local area networks and wide area networks.
Remember how earlier we linked ISDN technology with wide area networks? That's the real mission behind ISDN: linking users efficiently to wide area networks. But ISDN does not provide a satisfactory solution to the equation when we decide to factor in local area networks.
Suppose you are a terminal user on a LAN. Within the LAN, you can access a broad range of services, all at very high access speeds. But as soon as you venture outside the LAN, into the exchange network, things usually slow down considerably. Remote access into various information services is usually at a much slower pace than LAN speeds.
Uploading and downloading of large files takes more time than you prefer. Wouldn't it be better if you could move data across town, or across the country, at multi-megabit speeds?
LAN technology is well defined and cost-effective. Wide area network technology is likewise, and getting better. What's needed is a way to bridge the gap between the two.
And that's where a metropolitan area network fits in.
But wait, you say. We already have "metropolitan area networks" in place. They use fiber optic technology, provide enormous quantities of bandwidth, and are competitively priced.
Companies like Teleport Communications (Staten Island, N.Y.) and Metropolitan Fiber Systems (Oak Brook, Ill.) have been supplying high-speed fiber circuits for several years.
To which I reply, "Yes, you're right. These companies, and others, have indeed been supplying high-capacity fiber circuits in the local exchange area. But these have all been private lines."
I'm talking about switched connections, like dial tone.
Aha, now there's the twist--switched high-speed fiber connections, like dial tone. That's what makes MAN technology so exciting.
Imagine the possibilities: high-speed file transfer, video conferencing, integrated voice, data, text, video and others. All the things you've seen predicted now are really possible through MAN technology.
The press, as you might expect, has already begun flooding us with wave after wave of articles describing MANs. Several of the major vendors, such as AT&T, and the RBOCs in general, are pursuing MAN technology agressively. So once again technology throws us a curve. Just as we begin understanding and assimilating a new direction in communications, another one appears.
If that has become a frustrating part of your job as a telecomm professional, just remember it took us over 70 years to make the jump from basic electromechanical switching (e.g., direct control, step by step) to electronic switching.
Since that point in time, the pace has speeded up dramatically. Already we have several likely standards for MANs, supplied by the IEEE (Institute of Electrical and Electronics Engineers), ANSI (American National Standards Institute) and Bellcore (Bell Communications Research).
In the case of the first, IEEE specification 802.6 provides the structure for a powerful fiber-based ring network. What's exciting about 802.6 is that for the first time we have a data communication standard which will also address voice communication issues.
ANSI has been actively developing the Fiber Distributed Data Interface (FDDI) specification, which is rapidly gaining momentum in the industry.
Finally, Bellcore has developed a specification called Switched Multimegabit Data Service (SDMS), which provides a high-speed interface between various fiber networks. Bellcore has also developed the Sunchronous Optical Network (SONET), a transmission specification for fiber optic links between central offices.
Bellcore's SMDS recommendation also incorporates the IEEE 802.6 specification. Interesting linkages, indeed.
FDDI describes a 100 Mb/s fiber-based network that uses dual counter-rotating rings. The architecture is designed to dynamically reconfigure and restore itself when a physical break in the cable occurs.
The access methodology makes it possible to support both voice and data communications on the rings. In particular, voice does not have to be packetized to operate successfully on the network.
This opens many doors for users to truly integrate voice with data applications.
The specification for FDDI is just about completed and approved, which means a major rush of products (even more than are already on the market) can be expected before the end of this year.
The IEEE's 802.6 recommendation incorporates the concept of dual counter-rotating rings (or buses), the same as FDDI. However, it provides greatly improved access control for a large number of attached devices. Its design makes an 802.6 network almost totally disaster-proof, at least on a "single point of failure" basis.
Network access efficiency is based on a queuing algorithm that was developed and tested in Australia by a graduate student and his professor. A company called QPSX Ltd. (owned by Telecom Australia) was formed to further develop and market the technology.
The firm already has a U.S. subsidiary, which has transferred the technology to AT&T to help bring it to market.
The concept involves what is called a "distributed queue dual bus" or DQDB algorithm.
One analogy for describing how the concept works was offered recently by Bill Morgan, a well-known consultant who specializes in leading edge networks. The basic idea is this: Before you send a message, alert the network in advance of the request.
Imagine you're waiting for a train going south (this is one of the two network buses). The next time a northbound train arrives (using the other bus going in the opposite direction), you place a reservation request in that train. When the southbound train arrives, it has a seat (actually a time slot) ready and waiting.
The concept is based in part on a network access technique called a Cambridge Ring, which divides a ring into time slots. While that original concept worked fine if network usage was low, it became inefficient quickly when usage increased.
What appears to distinguish the DQDB algorithm from all others is that it provides the efficiency of an Ethernet at low access rates and similar efficiences to a Token Ring during peak demand periods.
In fact, according to Morgan, it is more efficient than either technique.
As defined by Bellcore, SMDS will be a "high-speed, connectionless, public packet switching service that will extend LAN-like performance beyond a subscriber's premises, across a metropolitan or wide area."
It will act, in some ways, similar to the way we use the switched telephone network to connect to other telephones.
However, its actual operation will be based as much on logical connections as physical linkages.
Already most of the RBOCs have announced SMDS programs.
We could see some trial SMDS products as early as this year.
Again, the role SMDS plays is that of a convenient, widely available method of connecting logical devices to other logical devices.
SONET is widely regarded as a major high-speed fiber transmission standard for connecting central offices.
This is critical for efficient utilization of fiber optics' vast transmission capabilities.
Much work has already been completed in SONET development, so much so that Northern Telecom recently announced a cadre of products supporting SONET.
Meaning To Users
So how are all these items related?
And what will it mean to you as a telecom professional?
These are the Next Generation of telecomm network services. They are already eclipsing ISDN in their potential.
Which of course begs the question: What becomes of ISDN? The answer is broadband ISDN...another story in itself.
These items are destined to be key elements in the network infrastructure of 1995, a structure based on lightwave technology.
Get familiar with them; they should become part of your vocabulary.
Now, when you hear someone talk about "the light at the end of the tunnel" you'll know how important that "light" really is.
How do all these new and exciting developments make the future look for you?
Very bright, indeed.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||metropolitan area networks|
|Date:||Apr 1, 1990|
|Previous Article:||User's idea brings trial of ISDN video.|
|Next Article:||Gateway saves on space, taxes.|