Given a noiseless protocol π0 computing a function f(x, y) of Alice and Bob’s private inputs x, y, the goal of interactive coding is to construct an error-resilient protocol π computing f such that even if some fraction of the communication is adversarially corrupted, both parties still learn f(x, y). Ideally, the resulting scheme π should be positive rate, computationally efficient, and achieve optimal error resilience.
Given a noiseless protocol π0 computing a function f(x, y) of Alice and Bob’s private inputs x, y, the goal of interactive coding is to construct an error-resilient protocol π computing f such that even if some fraction of the communication is adversarially corrupted, both parties still learn f(x, y). Ideally, the resulting scheme π should be positive rate, computationally efficient, and achieve optimal error resilience.
In interactive coding, Alice and Bob wish to compute some function f of their individual private inputs x and y. They do this by engaging in an interactive protocol to jointly compute f(x,y). The goal is to do this in an error-resilient way, such that even given some fraction of adversarial corruptions to the protocol, both parties still learn f(x,y). Typically, the error resilient protocols constructed by interactive coding schemes are \emph{non-adaptive}, that is, the length of the protocol as well as the speaker in each round is fixed beforehand. The maximal error resilience obtainable by non-adaptive schemes is now well understood. In order to circumvent known barriers and achieve higher error resilience, the work of Agrawal, Gelles, and Sahai (ISIT 2016) introduced to interactive coding the notion of \emph{adaptive} schemes, where the length of the protocol or the speaker order are no longer necessarily fixed. In this paper, we study the power of \emph{adaptive termination} in the context of the error resilience of interactive coding schemes. In other words, what is the power of schemes where Alice and Bob are allowed to disengage from the protocol early? We study this question in two contexts, both for the task of \emph{message exchange}, where the goal is to learn the other party's input.
How can we encode a communication protocol between two parties to become resilient to adversarial errors on the communication channel? If we encode each message in the communication protocol with a "good" error-correcting code (ECC), the error rate of the encoded protocol becomes poor (namely O(1/m) where m is the number of communication rounds). Towards addressing this issue, Schulman (FOCS'92, STOC'93) introduced the notion of interactive coding. We argue that whereas the method of separately encoding each message with an ECC ensures that the encoded protocol carries the same amount of information as the original protocol, this may no longer be the case if using interactive coding. In particular, the encoded protocol may completely leak a player's private input, even if it would remain secret in the original protocol. Towards addressing this problem, we introduce the notion of knowledge-preserving interactive coding, where the interactive coding protocol is required to preserve the "knowledge" transmitted in the original protocol. Our main results are as follows: The method of separately applying ECCs to each message has essentially optimal error rate: No knowledge-preserving interactive coding scheme can have an error rate of 1/m, where m is the number of rounds in the original protocol; If restricting to computationally-bounded (polynomial-time) adversaries, then assuming the existence of one-way functions (resp. sub exponentially-hard one-way functions), for every ϵ > 0, there exists a knowledge-preserving interactive coding schemes with constant error rate and information rate n-ϵ(resp. 1/polylog(n)) where n is the security parameter; additionally to achieve an error of even 1/m requires the existence of one-way functions; Finally, even if we restrict to computationally-bounded adversaries, knowledge-preserving interactive coding schemes with constant error rate can have an information rate of at most o(1 log n). This results applies even to non-constructive interactive coding schemes.
Abstract. Consider two parties Alice and Bob, who hold private inputs x and y, and wish to compute a function f(x, y) privately in the information theoretic sense; that is, each party should learn nothing beyond f(x, y). However, the communication channel available to them is noisy. This means that the channel can introduce errors in the transmission between the two parties. Moreover, the channel is adversarial in the sense that it knows the protocol that Alice and Bob are running, and maliciously introduces errors to disrupt the communication, subject to some bound on the total number of errors. A fundamental question in this setting is to design a protocol that remains private in the presence of large number of errors. If Alice and Bob are only interested in computing f(x, y) correctly, and not privately, then quite robust protocols are known that can tolerate a constant fraction of errors. However, none of these solutions is applicable in the setting of privacy, as they inherently leak information about the parties ’ inputs. This leads to the question whether we can simultaneously achieve privacy and error-resilience against a constant fraction of errors. We show that privacy and error-resilience are contradictory goals. In particular, we show that for every constant c> 0, there exists a function f which is privately computable in the error-less setting, but for which no private and correct protocol is resilient against a c-fraction of errors.
Consider two parties Alice and Bob, who hold private inputs x and y, and wish to compute a function f(x, y) privately in the information theoretic sense; that is, each party should learn nothing beyond f(x, y). However, the communication channel available to them is noisy. This means that the channel can introduce errors in the transmission between the two parties. Moreover, the channel is adversarial in the sense that it knows the protocol that Alice and Bob are running, and maliciously introduces errors to disrupt the communication, subject to some bound on the total number of errors. A fundamental question in this setting is to design a protocol that remains private in the presence of large number of errors.
Consider two parties, Alice and Bob, who hold private inputs x and y, and wish to compute a function f (x, y) privately in the information theoretic sense; that is, each party should learn nothing beyond f (x, y). However, the communication channel available to them is noisy. This means that the channel can introduce errors in the transmission between the two parties. Moreover, the channel is adversarial in the sense that it knows the protocol that Alice and Bob are running, and maliciously introduces errors to disrupt the communication, subject to some bound on the total number of errors. A fundamental question in this setting is to design a protocol that remains private in the presence of large number of errors. If Alice and Bob are only interested in computing f (x, y) correctly, and not privately, then quite robust protocols are known that can tolerate a constant fraction of errors. However, none of these solutions is applicable in the setting of privacy, as they inherently leak information about the parties' inputs. This leads to the question whether we can simultaneously achieve privacy and error-resilience against a constant fraction of errors. We show that privacy and errorresilience are contradictory goals. In particular, we show that for every constant c > 0, there exists a function f which is privately computable in the error-less setting, but for which no private and correct protocol is resilient against a c-fraction of errors.
We consider the problem of implementing two-party interactive quantum communication over noisy channels, a necessary endeavor if we wish to fully reap quantum advantages for communication. For an arbitrary protocol with n messages, designed for noiseless qudit channels (where d is arbitrary), our main result is a simulation method that fails with probability less than 2−Θ (nє) and uses a qudit channel n (1 + Θ (√є)) times, of which an є fraction can be corrupted adversarially. The simulation is thus capacity achieving to leading order, and we conjecture that it is optimal up to a constant factor in the √є term. Furthermore, the simulation is in a model that does not require pre-shared resources such as randomness or entanglement between the communicating parties. Perhaps surprisingly, this outperforms the best known overhead of 1 + O(√є loglog1/є) in the corresponding classical model, which is also conjectured to be optimal [Haeupler, FOCS’14]. Our work also improves over the best previously known quantum result where the overhead is a non-explicit large constant [Brassard et al., FOCS’14] for low є.