Volume Number: 15 (1999)
Issue Number: 10
Column Tag: Programmer's Bookshelf

Review: Cryptography and Network Security

by Paul E. Sevinc, Switzerland

Principles and Practice, 2nd edition

Cryptology is a topic more and more people in the IT business have to be familiar with. If you already know the basics and only need a reference manual so to speak, choosing the right book is easy: [Menezes et al. 1997]. But if you're looking for an introductory text book, you have an ordeal of choice. One book you'll often see recommended is [Schneier 1996]. In this article we're going to take a closer look at another one, namely the 2nd edition of Cryptography and Network Security: Principles and Practice [Stallings 1999].

Cryptography and Network Security (CaNS for short) has been published by Prentice Hall and was written by William Stallings, an author with quite an impressive publication list (see <http://www.prenhall.com/stallings/>). Besides, famous cryptographers like Xuejia Lai, Ron Rivest, Phil Zimmermann and many others reviewed Stallings' treatment of their special fields. In my opinion, this makes the book very trustworthy.

Between the introduction and the appendix (for teachers), CaNS is organized in four parts: Conventional Encryption, Public-Key Encryption and Hash Functions, Network Security Practice, and System Security. The introduction discusses network security models and computer security threats, thereby motivating the remainder of the text.

From the very beginning, it is obvious that CaNS is not intended to be a popular-science book but rather for (prospective) scientists and engineers.

The first part, chapters 2 to 5, deals with symmetric cryptography, i.e. schemes in which the same key is used for both encryption and decryption. Actually, the main focus is on so-called block ciphers while stream ciphers are hardly an issue. (There are two ways how to use block ciphers as stream ciphers, though, and Stallings explains both when he treats the four block-cipher modes of operation.) The math used in this part ñmodular arithmetic, linear equations, matricesñ should be basic to most programmers.

Chapters 2 & 3 treat the building blocks of both classical and modern symmetric algorithms and show how these are used within DES, the Data Encryption Standard. With the exception of the one-time pad, the classical algorithms are nowadays insecure, but still interesting for more than just historical reasons. Chapter 3 ends with general design principles of and attacks on block ciphers. I find this to be useful to practitioners who have to decide on and implement an algorithm, not to hackers :-)

Unless interested in details of different algorithms currently in use, the reader can safely skip most of chapter 4. Nevertheless, the first section is important because it contains information about the Men-in-the-Middle attack, and the last section is a good summary of the rest of the chapter.

Most of us don't really need to know how exactly encryption works. But if we use it, we better use it right. In chapter 5, Stallings explains where and how symmetric schemes can be used in an internetwork, including the non-trivial task of key distribution. One problem in this context is generating good pseudo-random numbers. Different approaches are mentioned, but only very briefly.

The second part, chapters 6 to 10, deals with functions that map variable-length data into a fixed-length value and with asymmetric cryptography, i.e. schemes in which different keys are used for encryption and decryption. This part is more mathematical than the last as public-key cryptosystems are heavily based on number theory, the subject of chapter 7. (Stallings only introduces what is necessary for understanding CaNS. If you're not familiar with number theory, I suggest you read this chapter before the 6th.)

Starting with the principles of public-key cryptography, chapter 6 presents RSA, the Rivest-Shamir-Adleman algorithm, the Diffie-Hellman key-exchange protocol, and key-management issues in general. At the end of this chapter, there's a short section about elliptic-curve cryptography, a hot topic of growing importance, which I highly appreciated.

Similar to chapters 3 & 4, chapter 8 first treats requirements and design principles of, as well as attacks on cryptographic hash functions and message authentication codes (key-dependent hash functions so to speak), and then chapter 9 describes four concrete algorithms in detail: MD5, SHA-1, RIPEMD-160 (that's what I call an acronym!), and HMAC.

Chapter 10 concludes the second part with a discussion of authentication protocols and digital signatures. The former include such based on symmetric cryptography, the latter DSS, the Digital Signature Standard.

The chapters in the third part (11-14) can be consulted in any order the reader likes. The prerequisites have been covered in the first two parts. This part is a nice add-on to older, in terms of IT years, computer-network books that don't cover network security (in enough detail). It includes sections on Kerberos, X.509, PGP, S/MIME, IPSec, SSL/TLS, and SET.

Kerberos and X.509 are for authentication purposes. PGP, Pretty Good Privacy, and S/MIME, Secure/Multipurpose Internet Mail Extensions, provide e-mail security. IPSec comprises the security features that have been defined for IPv6, but that can also be used with IPv4. (Those of you familiar with IP, the Internet Protocol, know that we're currently making the [slow] transition from IPv4 to IPv6.) SSL is Netscape's Secure Socket Layer which has become Transport Layer Security, an Internet standard. Finally, SET, Secure Electronic Transaction, is a specification initiated by MasterCard and Visa for credit-card payments over the internet.

By carefully reading several of these case studies, one gets a good idea of the trade-offs (e.g., security vs. complexity) involved in the design of cryptographic protocols. This is supported by Stallings' clear and concise writing and his judicious use of figures and tables.

The last part, chapters 15 & 16, is very high-level and easy to understand. However, it only provides an overview and can't replace an in-depth (and more technical) treatment of the same topics.

I was blown away by chapter 15. It is about intruders (i.e., hackers and crackers) and malicious programs, especially viruses. The many ñsometimes ingeniousñ ways of attacking a computer system and their countermeasures make for very exciting reading; look forward to a rainy Sunday!

Chapter 16 is a short chapter about the principles and goals underlying firewalls. It ends with a section on so-called trusted systems and how these can be used to defend against trojan horses.

CaNS is a very good introduction to cryptography, enabling you to understand and discuss the security of practical cryptosystems (e.g., the Mac OS keychain). But if you intend to develop or consult on cryptosystems yourself, you need to know more about number and information theory (see the seminal [Shannon 1948]) than is covered by Stallings. The parts about network and system security are good, too, even though they might not be applied enough for some readers (e.g., system administrators).

Every chapter ends with a set of non-trivial problems (the solutions are not part of the book). Some chapters additionally feature an appendix of their own, for example about ZIP's compression algorithm, the Birthday paradox, IPv4 and IPv6, etc. These appendices increase CaNS' usefulness and are thus worth being studied.

Finally, a five-page glossary and a one-page list of acronyms conclude this highly recommendable book.

I would like to thank Neso Atanasoski for his comments on this article and Andrew S. Downs for his editing of my Core Java review (May issue). Sorry for the delay, Andrew!

• [Menezes et al. 1997] MENEZES, Alfred J., VAN OORSCHOT Paul C., VANSTONE, Scott A. ñ Handbook of Applied Cryptography, CRC Press, 1997.
• [Schneier 1996] SCHNEIER, Bruce. ñ Applied Cryptography, Wiley, 2nd ed. 1996.
• [Shannon 1948] SHANNON, Claude E. ñ A Mathematical Theory of Communication, pp. 379-423 (623-656) in: The Bell System Technical Journal, volume 27, July (October) 1948.
• [Stallings 1999] STALLINGS, William. ñ Cryptography and Network Security: Principles and Practice, Prentice Hall, 2nd ed. 1999.

Paul is an EE student at the Swiss Federal Institute of Technology Zurich (ETHZ) where he might spend another three years if he accepts the Ph.D. position offered to him. But as he feels like living in the US for some time, he's wondering whether he should rather join a R&D department in Silicon Hills or Silicon Valley. While he's thinking about this, visit his homepage at http://www.stud.ee.ethz.ch/~psevinc/.