Re-Imagining Cyber Security

Tag: hypothesis

The Science of Intrusion Analysis and Incident Response: Introduction

[important]This is the first of several posts in a series expanding on how to turn intrusion analysis into a science.  Subscribe to the blog via email, follow us on twitter or like us on Facebook to keep-up!  [/important]

Previously I wrote about the Art of Intrusion Analysis and how I thought that Michelangelo’s quote was the best representation of how intrusion analysts arrive at knowledge.

However, my concern is not to document the art of Intrusion Analysis or Incident Response, but rather to transform the art into a science.  What does that mean?  What is the science of intrusion analysis and incident response?

First, we must define science (there are many definitions, this one will suffice for our purposes).

Science (from Latinscientia, meaning “knowledge”) is a systematic enterprise that builds and organizes knowledge in the form of testable explanations and predictions about the universe. — Wikipedia

Second, how will we know when intrusion analysis and incident response have become a science?  Aristotle can give us the answer.

[A] man knows a thing scientifically when he possesses a conviction arrived at in a certain way, and when the first principles on which that conviction rests are known to him with certainty—for unless he is more certain of his first principles than of the conclusion drawn from them he will only possess the knowledge in question accidentally.  — Aristotle (384 BC – 322 BC) in Nicomachean Ethics

From this I draw the following requirements to make intrusion analysis and incident response into a science:

  1. Intrusion Analysis and Incident Response must be systematic
  2. There must be first principles upon which hypotheses and predictions can be drawn and tested with experimentation
  3. There must be an organizing function to build knowledge
  4. There must be a set of theories which are generally accepted, testable, and repeatable following from first principles and hypotheses

Why do we care if Intrusion Analysis is a science or not?  An intrusion analysis and incident response science means less duplication of effort solving the same problems and a more cohesive approach to improving tools, tradecraft, and training.

Thanks to Richard (@taosecurity and Tao Security Blog) for the unanticipated use of his image! 🙂


20 Questions for an Intrusion Analyst

There are many approaches to finding the right people with the right talent to solve problems.  Intrusion analysis and incident response is no different.

I recently saw a great recruiting quiz to test potential employees in various knowledge areas which included programming, packet analysis, protocol analysis, snort rule writing, reverse engineering, data encoding, advanced mathematics, and other topics.  The test was designed so that it crossed so many topics one person would likely not successfully complete it.  However, it would highlight a person’s strengths and interests to give the assessor a more complete picture of the applicant.

This made me think, what topics and questions would I use to achieve the same effect?   After some deliberation, I have developed my own “20 Questions for an Intrusion Analyst” recruitment quiz (below) to highlight areas I think are important about a potential analyst joining a team.

As you may notice, I have covered several areas with these questions: analytic reasoning, creativity, adversary operations, packet analysis, intrusion detection, programming, reverse engineering, vulnerability analysis, exploit writing, and teaming.

I am purposefully not providing the answers 🙂

20 Questions for an Intrusion Analyst

  1. Describe you first experience with a computer or network threat
  2. You are given 500 pieces of straw and told that one piece is a needle which looks like straw.  How would you find the needle?  What other pieces of information would you like to have?
  3. Explain the difference between intrusion and extrusion detection
  4. Describe an adversary pivot, give an example, and explain its importance to intrusion analysis.
  5. Describe your analytic biases.
  6. Use the bit string 1101 to answer the following questions:
    1. The bit string when XORed with 0
    2. The decimal value of the string
    3. The string represented in hexadecimal
    4. Does this represent a printable ASCII character?  If so, which character?
  1. What is your favorite intrusion detection system?  What are its biases and limitations?
  2. Circle any of the following films you have seen: Hackers, War Games, Sneakers, Tron
  3. Describe a method to find an intruder using only network flow data (no content).
  4. Explain insertion and evasion of intrusion detection systems.  Give an example.
  5. Describe the activity detected by the following Snort rule.  What could be done to make the rule more effective?   alert icmp $EXTERNAL_NET any <> $HOME_NET any (msg: “activity alert!”; sid:10000011; content:”MZ”;)
  6. Write a code snippet to sort the following data by the first column
  1. How much time/week do you spend on your own researching computer security/threat topics?  What sources do you use to maintain situational awareness on threats in the wild?
  2. What will the following code print out?  Is there a vulnerability in the code?  If so, describe the vulnerability and a potential method of exploitation.
int main(int argc, char *argv[])
   char string[40];
   strcpy(string, argv[1]);
   printf("The message was: %s\n", string);
   printf("Program completed normally!\n\n");
   return 0;
  1. Describe and explain any “interesting” entries in the netstat log:
Proto Local Address     Foreign Address    State
  1. A host sends out an ICMP ECHO REPLY packet.  List all of your hypotheses to explain this activity.
  2. Describe the protocol stack of the following packet and the payload. Is the packet legitimate? Why or why not?
0000  00 00 c0 9f a0 97 00 a0 cc 3b bf fa 08 00 45 10   .........;....E.
0010  00 89 46 44 40 00 40 06 72 c7 c0 a8 00 02 c0 a8   ..FD@.@.r.......
0020  00 01 06 0e 00 17 99 c5 a1 54 17 f1 63 84 80 18   .........T..c...
0030  7d 78 cc 93 00 00 01 01 08 0a 00 9c 27 34 00 25   }x..........'4.%
0040  a6 2c ff fa 20 00 39 36 30 30 2c 39 36 30 30 ff   .,.. .9600,9600.
0050  f0 ff fa 23 00 62 61 6d 2e 7a 69 6e 67 2e 6f 72   ...#.bam.zing.or
0060  67 3a 30 2e 30 ff f0 ff fa 27 00 00 44 49 53 50   g:0.0....'..DISP
0070  4c 41 59 01 62 61 6d 2e 7a 69 6e 67 2e 6f 72 67
0080  3a 30 2e 30 ff f0 ff fa 18 00 78 74 65 72 6d 2d   :0.0......xterm-
0090  63 6f 6c 6f 72 ff f0                              color..
  1. What type of encoding is used in this example: aGVsbG8gd29ybGQNCg==
  2. Who do you turn to most on technical questions?

You didn’t expect the 20th question to be here did you?  You should expect the unexpected by now.

Beware Occam’s Razor

Occam’s Razor: A principle that generally recommends that, from among competing hypotheses, selecting the one that makes the fewest new assumptions usually provides the correct one, and that the simplest explanation will be the most plausible until evidence is presented to prove it false.  [Wikipedia]

Intrusion analysts face a unique problem: discovering and tracking adversaries not wanting to be discovered or tracked.  These adversaries will take significant action to prevent their operations from being discovered [see Insertion, Evasion, and Denial of Service: Eluding Network Intrusion Detection for examples].  Metasploit, a common exploitation framework, works hard to prevent their tool and modules from detection and signature including creating an Secure Sockets Layer (SSL) encrypted command-and-control channel.

In this environment, the cyber adversary is using Occam’s Razor against the analyst to elude detection.  Given a set of network traffic or events, an adversary wants any unintended observer to believe their traffic is normal and benign.  An adversary practices Denial and Deception in every packet.

Therefore, if an intrusion analyst relies too heavily on Occam’s Razor to describe a network event as benign, but a clever adversary is hiding their activities in that traffic – the adversary wins.

On the other hand, if an analyst does not employ Occam’s Razor effectively in their work, every packet will look suspicious wasting their precious time on unimportant events.

Richards Heuer’s preeminent work on the Psychology of Intelligence Analysis describes a very effective framework for employing the theory of Competing Hypothesis to decide which possible conclusion is the best given a set of facts and assumptions.  However, a common attack on Heuer’s framework is that an adversary utilizing Denial and Deception can easily defeat the analyst and have them select the conclusion the adversary wishes and not the one best describing the activity.

In the end, an intrusion analyst must use both a questioning and suspicious mind with a modicum of Occam’s Razor.

Powered by WordPress & Theme by Anders Norén