During the mid-1970s, computer enthusiasts were frequently as knowledgeable about the telephone system and telecommunications as they were about the details of their microprocessors. In fact, the details of telephone switching systems, cross-bar mechanisms, signaling, and other matters were common topics of discussion at Homebrew Computer Club meetings. (I remember once being accosted during "random access" by a rather famous fellow waving schematics related to an ESS switch!) Still, it's surprising that computer and telecommunications technologies continue to be viewed as immiscible disciplines, especially when you consider that they face many of the same problems and that their histories are so intertwined.
During the '70s, the architects of the Internet felt quite safe keeping the address space at a mere 32-bit size, never imagining this might be consumed before the end of the century. But the unanticipated and tremendous growth in computer networking in the '80s changed this. We now use large numbers of interacting computer systems, so that computer networking forms the basis of a modern information infrastructure. It's disquieting, though, that this revolution in information exchange has occurred removed from the world's largest connected network--the telephone system.
In Global Telecommunications: Layered Networks, Layered Services, Robert Heldman seeks to outline his vision for the integration of these two worlds from the telecommunications industry perspective. While the book fails in its stated goal of showing you how to take advantage of communications technology (a daunting task for any single book), it does provide what may be a more valuable service: Global Telecommunications lays bare the frustrations of bringing about the birth of this global information industry. In a rambling style, the author leads us through the process by which the telecom industry structures the problem of a unified voice, video, and data-communications network. That the book is targeted at the telecommunications industry is apparent in its use of telecom terminology and approaches, and its attempts to span the gulf between the nuts-and-bolts physical realities and the top-down market-driven business analysis of a telecommunications corporate participant.
This approach normally would be guaranteed to annoy both the dyed-in-the-wool computer user and the telecommunications professional. Computer professionals quickly become frustrated at being sold a "multimedia soon" future which requires megabit (perhaps gigabit) service, while having to put up with current piddling kilobit-per-second analog modem communications. On the other hand, any astute telecommunications professional is already aware of the difficulty of implementing even the most mundane of ideas presented in this book. However, even though it oscillates between microscopic details (bit-density problems with T1) and cosmic issues (Maslow's Hierarchy of Needs), Global Telecommunications provides us with a unique view, a "gestalt" if you will, of an industry at the crossroads. In fact, after a while I began to feel that I was reading a psychoanalytic profile of the telecommunications industry in crisis. It underscored how difficult it will be for the industry to take full advantage of the new information age.
For example, the book's perspective on data-communications networking (page 123) is quite revealing: "We are expanding from the traditional analog voice-based network, with voice-only services...we are beginning to use digital capabilities to better operate and expand the network."
In essence, data is viewed as a special case of voice! This isn't really surprising, since that's what the telecommunications industry made its fortune on. However, this is exactly opposed to the computer industry's approach, where digital information is everything. Every media--voice, data, video--is digital.
Another example of the book's somewhat myopic view can be found in its treatment of ISDN. Like fusion, ISDN has been "just around the corner" for at least a decade. While its genesis occurred at about the same time the first PCs appeared, it's been slow going for ISDN ever since. The problem is that the telcos don't view ISDN as an extension of LANs, but as a replacement for them (pages 147-148).
By delaying the public network offering in the 1970s and 1980s, the RBOCs [regional Bell Operating Companies] have driven the MIS managers of large firms to create new networks that are quite autonomous from the telcos.... This lack of success encouraged considerable growth of point-to-point LAN-type architectures.... This has resulted in a separate "private networking" approach to interconnect LANs by gateways, bridges and routers. However, expansive growth has made this cumbersome, quite expensive and somewhat limited to closed user groups. In the new world of multiple systems interfacing together, there is a need for the more robust public networks to carry this traffic to and from the internal private networks.
In fact, LANs have been fantastically successful, partially due to their relatively low cost and availability, and partially to their appropriateness in treating data like data, not a special case of voice. It's ludicrous to believe that ISDN will recapture the bandwidth lost to private data networks over the next decade without examining issues such as performance requirements and existing software interfaces.
However, as pointed out in Global Telecommunications, digital communications still amount to only 3 percent of the total current service present, so you can expect voice to be in the driver's seat over the long term (the next 40 years?) as the industry develops new equipment plans. Quite simply, until the telcos see an economic necessity, they will not take data seriously. (Given the rapid rate of change in this area, they could end up missing the boat completely.)
A point not addressed in Heldman's book is "why" the telephone system went digital in the first place. Steve Hardwick shows in ISDN Design that it was the economics of digital transmission, switching, and signaling that caused the industry to migrate incrementally away from ancient analog mechanisms. This has had the indirect result of tossing the entire telecommunications industry into the information market, since now voice is just one of many kinds of information to be transmitted. This was not anticipated, but rather dictated by circumstance.
In a computer-oriented world, where entire computer systems become obsolete after a few years, there's little patience for ISDN, even though it fundamentally extends the reach of computer technology to any point in the world and has the potential for turning every telephone into a computer system. Compared to the computer industry's hare, the conservative, button-down telecommunications industry begins to look like a tortoise by always attempting to justify each step before making it.
The irony of this is that to be in the "information side" of the business, you must be immersed in the application of information. Unfortunately, the monolithic firms capable of orchestrating gigantic phone networks cannot seem to grasp this simple fact.
Global Telecommunications is valuable for its isolated vignettes focusing on the needs of the telecommunications customer as well as outlining (sometimes futile) attempts to provide what the customer desired. It is worthwhile to compare this approach with the breakneck pace of high-speed computer networking (like you find within the Internet community). Of marked contrast is the mind-boggling complexity of the low level of abstraction present in telecommunications (mostly at the ISO physical and link layers) versus the relatively trivial components in most computer-networking fields. (I used to have an entire bookshelf dedicated just to a partial description of ISDN, while a single book on TCP/IP was sufficient.) It is this holdover from the past that forces the tortoise carrying the world on its back to pale beside the lightweight hare that uses abstraction to minimize burdens. The tortoise has a problem in scaling down; the hare has a similar problem in scaling up. One measures time in decades, while the other is unwilling to wait for a fraction of a year for fear of missing a technology window.
Global Telecommunications' unstructured "all over the map" approach is its worst characteristic. You just can't read it in a typically linear fashion and expect to keep a logical train of thought. Another irritation is the frequent mention of market-driven customer-application approaches that are completely naive regarding digital-communications applications. (There is no mention, for example, of remote file systems, distributed computing, or interoperability.) In this, the hallmarks of modern computer-networking progress of recent years are completely ignored. Alas, this is a major omission in an industry that prides itself on its technical competence and thoroughness.
Telecommunications is rapidly approaching a crossroads. Recent court and regulatory decisions mean that other firms can access telephone subscribers directly, as an alternative to local exchange carriers (LECs). Many powerful interests are already jockeying for position to compete with them on an equal basis to provide even more services than the existing LECs provide. Global Telecommunications correctly puts forth a number of barriers for the industry to hurdle, but it misses the most significant one: employing the existing computer networking technology to accelerate the demand and acceptance of modern communications technologies. In other words, the compelling need for new communications technology is rooted in developing the existing information-system paradigm, not in trying to invent something totally new or recast things in a past image.
There may be some historical significance to Global Telecommunications in that it captures a moment in time before the pent-up innovation of one industry impacts another, very much like the old saw of an irresistible force meeting an immovable object. I'm sure it will be an interesting sight, regardless of the outcome.
Copyright © 1992, Dr. Dobb's Journal