BKMLCRPT.RVW   20041012

"Malicious Cryptography", Adam L. Young/Moti Yung, 2004,
0-7645-4975-8, U$45.00/C$64.99/UK#29.99
%A   Adam L. Young
%A   Moti Yung
%C   5353 Dundas Street West, 4th Floor, Etobicoke, ON   M9B 6H8
%D   2004
%G   0-7645-4975-8
%I   John Wiley & Sons, Inc.
%O   U$45.00/C$64.99/UK#29.99 416-236-4433 fax: 416-236-4448
%O  http://www.amazon.com/exec/obidos/ASIN/0764549758/robsladesinterne
  http://www.amazon.co.uk/exec/obidos/ASIN/0764549758/robsladesinte-21
%O   http://www.amazon.ca/exec/obidos/ASIN/0764549758/robsladesin03-20
%P   392 p.
%T   "Malicious Cryptography: Exposing Cryptovirology"

Both the foreword and the introduction are turgid, and bloated with
excessive verbiage, while never giving a clear indication of what the
book is actually about.  Does it have to do with viruses at all?  Is
it about the use of cryptography in any kind of criminal or unethical
endeavour?  The initial material does not make this clear. 
Occasionally the text becomes so flowery that sentences have no
meaning at all.

The lack of clarity is not assisted by the creation of new and
idiosyncratic terms, or the use of existing jargon in non-standard
ways.  In chapter one, a fictional and glacially slow trip through the
mind of a virus writer, we are told that self-checking modules that
some programs use to detect modification in their own code are
"beneficial Trojans" or "battleprogs."  The term multipartite is
defined in such a way that merely copying the program into RAM (Random
Access Memory) qualifies: that would make every virus ever written,
and every program, for that matter, multipartite.  "Kleptogram" is
used throughout the book, but only defined (and not very clearly) in
the last chapter.  Releasing any virus is seen as having something to
do with "information warfare," which would agree with many
sensationalistic journalists who have written on the subject, but
would probably surprise legitimate experts such as Dorothy Denning. 
"Virology" itself (and the more specialized "cryptovirology") is an
excellent term for computer virus research--it just isn't used very
widely.  There is a glossary: it defines commonly known terms and does
not define the specialized jargon that the authors have used.

The confusion is not limited to terminology.  There is no technical
sense to the statement (on page twenty five) that a certain layer of
the network stack is "high enough to facilitate rapid software
development" (compilers don't care where their software ends up) but
low enough to escape detection (files, processes, and network packets
are all visible).  A disk locking program, as described, would have no
effect on the operations of a remote access trojan.  And, of course,
our fictional protagonist is constantly creating new versions of the
mythical "undetectable" virus, without there being any indication of
how this might be done.

(The fictional aspects of the book are not limited to chapter one. 
Throughout the work, examples are taken from fiction: it certainly
feels like more illustrations come from works like "Shockwave Rider"
and "Alien" than from real life.)

Chapter two starts to get a bit better.  The authors introduce the
idea of using asymmetric cryptography in order to create a virus (or
other piece of malware) that, rather than merely destroying data,
provides for a reversible denial of access to data, and therefore the
possibility of extortion.  The idea is academically interesting, but
there might be a few practical details to be worked out.

Chapter three seems to move further into the academic realm, with an
interesting overview of issues in regard to the generation of random,
or pseudorandom, numbers.  There is also an initial exploration of
anonymity, with an insufficient description of "mix networks" (onion
routing being one example).  A little more discussion of anonymity
starts off chapter four, which then moves on to another use of
asymmetric cryptography in malware: the "deniable" recovery of stolen
information, via distribution over public channels.  Cryptocounters,
which could be used to store generational or other information about
the spread of a virus, without such data being accessible to virus
researchers, are discussed in chapter five.  Chapter six looks at
aspects of searching for, and retrieving, information without
disclosing the fact that an exploration is occurring.  However, much
of the material appears to be some highly abstract solutions rather
desperately in search of problems.  Varying the extortion scenario,
chapter seven proposes a viral network that could retaliate for
disinfection of any node by threatening disclosure of sensitive
information.  While the analysis of the structure of the attack is
sound, the assumption of payoffs, coercion, and undetectability leave
something to be desired.

Chapter eight examines the standard antiviral processes (signature
scanning, activity monitoring, and change detection) with some
miscellaneous explorations, although the discussion is prejudiced by
the assumption that we are dealing with traditional (and no longer
widely used) file infectors.  Trojan horse programs are not terribly
well defined in chapter nine.  (I was amused at the disclaimer given
when the issue of "salami" scams was raised: I have found reliable
evidence for only one, extremely minor, instance of the device.) 
Subliminal channels are means of passing information via cryptographic
keys, but chapter ten is not very clear in regard to their use. 
SETUPs (Secretly Embedded Trapdoor with Universal Protection) are
discussed in chapter eleven, although the authors appear to admit that
this is only an academic exercise: there are easier attacks.  Another
form is discussed in chapter twelve.

Does this book fulfill its function?  That rather depends on what the
intent of the work was, which is far from clear.  Was the text
intended to be a reference for some interesting topics in
cryptography?  The verbiage and lack of structure would be a
difficulty for those seeking to use it so.  Is the publication
directed at the general public?  The audience of those who read number
theoretical manuscripts for fun might be a bit limited.  (I've got to
say that "Algebraic Aspects of Cryptography" [cf. BKALASCR.RVW] was an
easier read, and it makes no pretence of being other than an
scholastic paper.)

Is the volume supposed to be a serious warning against new forms of
malware?  The inclusion of a great deal of extraneous content and the
lack of clear explanations or examples of some basic concepts limit
the value of the work in this regard.  In addition, much of the
material concentrates on building more malign malware, rather than
dealing with defence against it.  (I'm not too worried about vxers
getting ideas from Young and Yung: implementing crypto properly is a
painstaking task, and from almost twenty years experience of studying
blackhat products and authors, I'm fairly sure there'd be lots of bugs
in what might be released.  On the other hand, somebody in a
government office might be working on Magic Lantern version 3.01 ...)

For those seriously involved in the study of viruses and malware this
book has some interesting points that should be examined, but little
of practical use.  For ardent students of cryptography, the work notes
some interesting areas of work.  For those seeking examples of writing
styles to emulate, please look elsewhere.

copyright Robert M. Slade, 2004   BKMLCRPT.RVW   20041012