Computing and Communication

Madhu Sudan (Microsoft & MIT)

Despite the fact that computers and communication devices are 
highly integrated today, the theories remain separate. The result 
is a weak theory and ad-hoc practice. Existing devices end up 
being unreliable, vulnerable to attack, and inefficient due to 
the adhoc nature of the practice. We propose that there should 
be a unified theory that models "computers that communicate", 
articulates the problems they wish to solve, and provides 
qualitative and quantitative metrics to evaluate and compare 
different solutions, thereby bringing some level of rigor into 
the practice.

The brief description above is meant to open a conversation. 
Some of the concerns are brought out in my talks on Semantic Communication 
(see e.g. http://people.csail.mit.edu/madhu/talks/2010/Semantic-UCSD.ppsx ), 
and I will be happy to explain further, or hear others thoughts on this theme.