Discussion Questions for Week 8:

1. The description of a "Revised Key Exchange Protocol"
on pp. 506-507 is internally inconsistent, so it just
doesn't make sense.  How might the description be fixed?  
Does your fixed version prevent a man-in-the-middle attack,
as the text claims?  If not, can you add further assumptions
about the attack model that would make the protocol useful?

2. Consider a communication system where messages are 
spoken out load (for example, a lecture or reading 
to a classroom). Invent a covert channel mechanism 
suitable for communicating a single secret number between 
1 and 9 (inclusive) within an arbitrary spoken message.  
(You may assume messages are fairly long --- say 
at least a few hundred words.) A good mechanism 
should reliably communicate the secret to listeners
who know the encoding, while other listeners cannot
even detect that any information is being passed.