Networks Timothy Ramteke ISBN 0-13-958059-X 1994, Prentice Hall Reviewed by: D. Crocker / Brandenburg Consulting August 24, 1995 As the field of networking has grown, rapidly, broadly and deeply, survey books have become increasingly difficult to write and essential to read. More newcomers need the introduction and more old-timers need to have at least passing familiarity with concepts and detail peripherally related to their daily work. So I approached "Networks" with considerable enthusiasm. A quick survey of the content I could evaluate easily left me confused. Eventually, I DID note that the title was "Networks" and not "Networking". This is a book primarily about telecommunications technology and not data networking. The author thinks otherwise and even suggests ways to use the book for WAN or LAN courses. So, here's the bad news and then I'll suggest why I'm glad to have the book around: "Networks" simply does not cover datagram-based networking very well. For a book published in 1994, it's odd that the author was unaware that DARPA hasn't been a major funder of the Internet for many years, makes no mention of NSFNet, and doesn't know that the Arpanet ceased to exist in 1989. The discussion on TCP acknowledgment and windowing is a good effort, but doesn't quite work. For a book introducing technical concepts, it's also odd that there are quite a few pages discussing Unix commands for Internet activity. The chapter on Netware also shows user interactions. The chapter called "Interconnecting LANs" combines bridging and routing with discussion of the newer, high-speed transmission technologies (e.g., frame relay and ATM.) The author cites assistance he received from the eminent Radia Perlman and the discussion on bridging and routing is, in fact, OK. Just OK. Don't design any backbones with this as your only reference material. Now the good news. The book starts with a section on history, dating back to 1831. The chapter on "Analog and Digital Signals" opens with a description of Ohm's Law! Various telephone transmission and switching technologies, e.g., Centrex and Signalling System 7, are covered as are several national telephone service architectures. This is fun stuff and quite eclectic. It tends to be primarily expository description of functional characteristics. Underlying detail and theory is often lacking, as is critical analysis. But the compendium of detail was easy to read and satisfied some long-standing curiosity I've had about the part of the world that is activated when I pick up a handset. Part I, Fundamentals, is 50 pages covering history, abstract architectures, basic signals and transmission technologies for the signals to travel over. In case you don't have an OSI 7-layered picture around, the one in "Networks" shows X.25 at the network layer, with no discussion of the layer's subdivision into components for internetworking. The application layer exemplars are, of course, X.400 and X.500. In other words, the author attempts to carefully tow the party line in the book's chapters. If you require clever personal asides ranging from soap boxes pronouncements to flaming criticism, this book will leave you unfulfilled. If you want simple and clear exposition, this book is quite comfortable. Part II, Voice Networking, is 155 pages, with chapters covering signaling and switching, public and private networks, call processing mechanisms, T1- speed systems and virtual networking. The 110 pages of Part III, Wide Area Networks, covers SNA, X.25, SS7, ISDN and SONET. So even in this section the telephone company focus persists. The final part is on LANs and Internetworking. Its 140 pages span LANS -- primarily access methods -- Netware, interconnection, and TCP/IP, as discussed above. So who SHOULD read this book? I'd recommend it for a datacomm person with no background on the phone technology. Same for programmers light on the underpinnings for communications infrastructure.