Search in Book Reviews

The book is rather short at about 170 pages and really only offers an overview. Though the book does cover all that is needed in 9 chapters. There are no circuit diagrams or source code in this book: it is at a much higher level. There is a good description of the OSI stack and diagrams of various processes e.g. TCP/IP implementation on Unix. There is a case study of a 2-layer switch. These are multi-board and multi-processor systems. If you are looking for more information on the nuts and bolts of how the Internet, national (digital) phone systems or WANs work then this is a good book.

Managers will find this book useful as it does explain how and why many choices are made and other pros and cons. For example the make verses buy arguments for some software modules. The relative merits for COTS RTOS over your own. Yes, in high-end communications software where 20 year MTBF is required, many still roll their own OS. Other topics discussed are things like HW redundancy schemes, hot and warm swap, fault tolerance etc.

One area I found interesting was the tools and debug. What was there was good but again it was rather short and superficial. Not really good enough if, as a manager, you are going to sign off in excess of 10,000 GB pounds (15K USD) on test gear for a project.

I think as an introduction it works but you will need other books to go any deeper, certainly for any implementation. I have free company specific and tool guides to networking with as much in as this book. However, it is well written and the author clearly understands the field. The author may want to follow up this book with some more (larger) volumes on specific areas that really can be used by engineers. Possibly a much enlarged second edition of the current book.

I think this would be useful for students and managers and possibly an introduction for engineers new to high-end communications systems.