The subject of this work is the possibility of private distributed computations of $n$-argument functions defined over the integers. A function $f$ is $t$-private if there exists a protocol for computing $f$, so that no coalition of at most $t$ participants can infer any additional information from the execution of the protocol. It is known that over finite domains, every function can be computed $\left\lfloor{(n-1)/2}\right\rfloor$-privately. Some functions, like addition, are even $n$-private.
We prove that this result cannot be extended to infinite domains. The possibility of privately computing $f$ is shown to be closely related to the communication complexity of $f$. Using this relation, we show, for example, that $n$-argument addition is $\left\lfloor{(n-1)/2}\right\rfloor$-private over the nonnegative integers, but not even $1$-private over all the integers.
Finally, a complete characterization of $t$-private Boolean functions over countable domains is given. A Boolean function is $1$-private if and only if its communication complexity is bounded. This characterization enables us to prove that every Boolean function falls into one of the following three categories: It is either $n$-private, $\left\lfloor{(n-1)/2}\right\rfloor$-private but not $\left\lceil{n/2}\right\rceil $-private, or not $1$-private.
[1]
G. R. Blakley,et al.
Security Proofs for Information Protection Systems
,
1981,
1981 IEEE Symposium on Security and Privacy.
[2]
Reuven Bar-Yehuda,et al.
Privacy, additional information and communication
,
1993,
IEEE Trans. Inf. Theory.
[3]
Eyal Kushilevitz,et al.
Privacy and communication complexity
,
1989,
30th Annual Symposium on Foundations of Computer Science.
[4]
Donald Beaver.
Perfect Privacy For Two-Party Protocols
,
1989,
Distributed Computing And Cryptography.
[5]
Josh Benaloh,et al.
Secret sharing homomorphisms: keeping shares of a secret secret
,
1987,
CRYPTO 1987.
[6]
Richard J. Lipton,et al.
Multi-party protocols
,
1983,
STOC.
[7]
Silvio Micali,et al.
How to play ANY mental game
,
1987,
STOC.
[8]
Avi Wigderson,et al.
Completeness theorems for non-cryptographic fault-tolerant distributed computation
,
1988,
STOC '88.
[9]
Judit Bar-Ilan,et al.
Non-cryptographic fault-tolerant computing in constant number of rounds of interaction
,
1989,
PODC '89.