Limitations of Quantum Advice and One-Way Communication

Scott Aaronson

DOI: https://doi.org/10.4086/toc.2005.v001a001

2018-06-22

Abstract:Although a quantum state requires exponentially many classical bits to describe, the laws of quantum mechanics impose severe restrictions on how that state can be accessed. This paper shows in three settings that quantum messages have only limited advantages over classical ones.

Quantum Physics,Computational Complexity

What problem does this paper attempt to address?

The problems that this paper attempts to solve mainly focus on the limitations of quantum advice and one - way communication. Specifically, the author Scott Aaronson explores three main issues:

1. **The relationship between quantum advice and classical advice**: The author studies whether states with quantum advice can be simulated by classical advice in polynomial time. In particular, he proves that \( \text{BQP}/\text{qpoly} \subseteq \text{PP}/\text{poly} \), that is, given quantum advice of polynomial size, problems solvable in quantum polynomial time can be simulated with classical advice in probabilistic polynomial time. This result solves a problem proposed by Buhrman and shows that we cannot expect to distinguish quantum advice from classical advice in the non - relativized case.

2. **The relationship between NP and BQP with quantum advice**: The author constructs a situation relative to an oracle such that \( \text{NP} \not\subseteq \text{BQP}/\text{qpoly} \). To this end, he provides the first correct proof of the quantum search direct product theorem. This theorem states that if a quantum algorithm makes fewer than \( o(\sqrt{N}) \) queries to an \( N \)-item database containing \( K \) marked items, the probability of finding all \( K \) marked items decreases exponentially with \( K \). This result not only corrects Klauck's previous incorrect proof but also provides new insights into other problems.

3. **Lower bounds on the quantum one - way communication complexity**: The author introduces a new trace distance method for proving lower bounds on the quantum one - way communication complexity. Using this method, he provides optimal quantum lower bounds for two problems proposed by Ambainis, for which there were not even known non - trivial classical random protocol lower bounds before. These results not only show the limitations of the VC - dimension method but also provide new tools for further studying the relationship between random and quantum one - way communication complexity.

In summary, the core issues of this paper are to explore the capabilities and limitations of quantum advice and one - way communication, and to reveal the potential limitations of quantum computing in certain cases through theoretical analysis and construction proofs.