Cryptography Reference
In-Depth Information
Chapter 18
Secure Multiparty Computation
In this chapter, we address secure multiparty computation (MPC) (i.e., the problem
of how mutually distrusting parties can compute a function without revealing their
individual input values to each other). We introduce the topic in Section 18.1, sum-
marize the major results in Section 18.2, and conclude with some final remarks in
Section 18.3. This chapter is intentionally kept short. Oded Goldreich has written
a primer entitled
Secure Multi-Party Computation
1
that is comprehensive and rec-
ommended reading for anybody interested or working in this area of research (the
material is also included in Chapter 7 of [1]).
18.1
INTRODUCTION
From a very high level of abstraction, almost all (cryptographic) problems can be
solved by specifying and actually implementing a random process that maps
n
inputs
to
n
outputs. The inputs to the process can be thought of as local inputs of
n
parties,
whereas the
n
outputs can be thought of as their local outputs. In either case, it has
to be distinguished whether there is a(n inside or outside) party that is trusted by all
parties.
•
If such a (trusted) party exists, then it can be used to implement the process.
In this case, the
n
parties can send their local inputs to the trusted party, and
the trusted party can then compute the outcome of the process and send each
party its local output. It goes without saying that all communications between
the
n
parties and the trusted party must take place over secure channels.
1
http://www.wisdom.weizmann.ac.il/
∼
oded/pp.html
