Information Security. Mark Stamp

cryptography, as well as issues with the protocols themselves. These issues make both of these topics interesting case studies.

      1.3.4 Software

      In the final part of the book, we'll take a look at some aspects of security that are specifically related to software. This is a huge topic, yet the two chapters in this book manage to hit on most of the fundamental issues. For starters, we'll discuss security flaws and malware, which were mentioned above. We'll also consider software reverse engineering, which illustrates how a dedicated attacker can deconstruct software, even without access to the source code.

1.4 The People Problem

Users are surprisingly capable when it comes to unintentionally inflicting damage on security systems. For example, suppose that Bob wants to purchase an item from, say, amazon.com . Bob can use his Web browser to securely contact Amazon using the SSL protocol (discussed in Part III), which relies on various cryptographic techniques (see Part I). Access control issues arise in such a transaction (Part II), and all of these security mechanisms are enforced in software (Part IV). So far, so good. However, we'll see that there is a practical attack on this transaction that Trudy can conduct, which will cause Bob's Web browser to issue a warning. If Bob heeds the warning, Trudy's attack will be foiled. Unfortunately, the odds are good that Bob will ignore the warning, which has the effect of negating this sophisticated security architecture. That is, the security can be broken due to user error, even if the cryptography, protocols, access control, and software all performed flawlessly.

      To take just one more example, consider passwords. Users want to choose easy to remember passwords, but this also makes it easier for Trudy to guess passwords. A possible solution is to assign strong passwords to users. However, this is generally a bad idea since it is likely to result in passwords being written down and posted in prominent locations, likely making the system less secure than if users were allowed to choose their own (weaker) passwords.

      As mentioned above, the primary focus of this book is on understanding security mechanisms—the nuts and bolts of security. Yet in several places throughout the book, various “people problems” arise. It would be possible to write several volumes on this topic, but the bottom line is that, from a security perspective, we would like to remove humans from the equation as much as is humanly possible.

      For more information on the role that humans play in information security, a good source is Ross Anderson's book [3]. Anderson's book is filled with case studies of security failures, many—if not most—of which have at least one of their roots somewhere in the actions of the supposed good guys, Alice and Bob. While we expect Trudy to do bad things, surprisingly often the actions of Alice and Bob serve to help, rather than hinder, Trudy.

1.5 Principles and Practice

This book is not a theory book. While theory certainly has its place, in your opinionated author's opinion, many aspects of information security are not yet ripe for a meaningful theoretical treatment.⁷ Of course, some topics are inherently more theoretical than others. But even relatively theoretical security topics can be learned to a reasonable depth without diving too deeply into the theory. For example, cryptography can be (and often is) taught from a highly mathematical perspective. However, with rare exception, a little elementary math is all that is needed to understand cryptographic principles.

      This book is certainly not an attacker's how‐to guide either. Nevertheless, your practical author has consciously tried to keep the focus on real‐world issues, but at a deep enough level to give the reader some understanding of—and appreciation for—the underlying concepts. The goal is to get into some depth without overwhelming the reader with excessive trivial details. Admittedly, this is a delicate balancing act and, no doubt, many will disagree that a proper balance has been struck. In your defensive author's defense, it should be noted that this book touches on a very large number of security issues related to a wide variety of fundamental principles. This breadth necessarily comes at the expense of some rigor and detail.

      For those who yearn for a more theoretical treatment of the some of the topics covered here, Bishop's book [10] is the obvious choice. There are numerous fine books and articles available that focus in more detail on the various security topics discussed in this book. Your favorite search engine will quickly reveal many such sources.

1.6 Problems

       The problem is not that there are problems. The problem is expecting otherwise and thinking that having problems is a problem.

      —Theodore I. Rubin

      1 Among the fundamental challenges in information security are confidentiality, integrity, and availability, or CIA.Define each of the terms confidentiality, integrity, and availability.Give a concrete example where both confidentiality and integrity are critically important.Give a concrete example where integrity is more important than confidentiality.Give a concrete example where availability is the overriding concern.

      2 From a bank's perspective, which is likely to be more important (and why), the integrity of its customer's data or the confidentiality of the data? From the perspective of the bank's customers, which is more important (and why)?

      3 Some authors distinguish between secrecy, privacy, and confidentiality. In this usage, secrecy is equivalent to our use of the term confidentiality, whereas privacy is secrecy applied to personal data, and confidentiality (in this misguided sense) is somewhat more restrictive than the terminology as used in this book, as it refers to an obligation not to divulge certain information.Discuss a real‐world situation where privacy is an important security issue.Discuss a real‐world situation where confidentiality (in this restricted sense) is a critical security issue.

      4 Cryptography is sometimes said to be “brittle,” in the sense that it can be very strong, but when it breaks, it's strength is shattered.⁸ In contrast, some security features can “bend” without breaking completely—security may be lost as a result of such bending, but some useful level of security can remain.Other than cryptography, give an example of a security mechanism that is brittle.Provide an example of a security mechanism that is not brittle, that is, the security can bend without completely breaking.

      5 Read Diffie and Hellman's classic paper [30].Briefly summarize the paper.Diffie and Hellman give a system for distributing keys over an insecure channel (see Section 3 of the paper). How does this system work?Diffie and Hellman also conjecture that a “one way compiler” might be used to construct a public key cryptosystem. Do you believe this is a plausible approach? Why or why not?

      6 The most famous cipher of World War II is the German Enigma. This cipher was broken by the Allies and intelligence gained from Enigma messages proved invaluable. At first, the Allies were very careful when using the information gained from broken Enigma messages—sometimes the Allies did not use information that could have given them an advantage. However, later in the war, the Allies (and, in particular, the Americans) were much less careful, as they tended to use virtually all information obtained from broken Enigma messages.Briefly discuss a significant World War II event where broken Enigma messages played a major role.The Allies were cautious about using information gained from broken Enigma messages for fear that the Germans would realize their cipher was compromised. Discuss two different approaches that the Germans might have taken if they had realized that the Enigma was broken.At some point, it should have become obvious to the Germans that the Enigma was broken, yet the cipher was used until the end of the war. Why did the Nazis continue to use the Enigma?

      7 When you want to authenticate yourself to your computer, most likely you type in your username and password. The username is considered public knowledge, so it is the password that authenticates you. Your password is “something you know.”It is also possible to authenticate based on “something you are,” that is, a physical characteristic. Such a characteristic is known as a biometric.