Verification of Universal Quantum Computation: Difference between revisions

From Quantum Protocol Zoo
Jump to navigation Jump to search
 
(40 intermediate revisions by the same user not shown)
Line 1: Line 1:
==Functionality==
==Functionality==
Quantum Computers perform task which are intractable for classical computers. The basic question here would be, "How should one verify the result of a quantum computer? This task is known as quantum verification or verification of quantum computation. '''Tags:'''  [[:Category: Quantum Functionality|Quantum Functionality]], [[Category: Quantum Functionality]] [[:Category:Universal Task|Universal Task]][[Category:Universal Task]]
Quantum Computers perform task which are intractable for classical computers. The basic question here would be, "How should one verify the result of a quantum computer? This task is known as quantum verification or verification of quantum computation. [https://complexityzoo.uwaterloo.ca/Complexity_Zoo:B#bqp BQP] is the class of problems that can be solved by a quantum computer and [https://complexityzoo.uwaterloo.ca/Complexity_Zoo:B#bpp BPP] is the class of problems that can be solved by a classical computer. BPP is contained in BQP and hence, there are problems a quantum computer would solve that are intractable for a classical computer, to put it simply. Thus, if in future, an untrusted company claims to have built a quantum computer, how can the consumer be sure of the [https://people.eecs.berkeley.edu/~sanjamg/classes/cs276-fall14/scribe/lec09.pdf correctness] of the results when he/she (the consumer) cannot compare the results predicted by the proposed quantum computer? This problem is addressed by the functionality, 'verification of quantum computers'.  Verification of universal quantum computation targets every computation that can be performed by a quantum computer.<br/><br/>
'''Tags:'''  [[:Category: Quantum Functionality|Quantum Functionality]], [[Category: Quantum Functionality]] [[:Category:Universal Task|Universal Task]][[Category:Universal Task]], [[Classical Verification of Universal Quantum Computation]], [[Verification of Sub-Universal Quantum Computation]], [[Verification of NP-complete problems]]


==Use Case==
* Quantum task
* Classical analogue: [[Classical Verification of Quantum Computation]]
* Best Implementation specifications
==Protocols==
==Protocols==
#Single-prover prepare-and-send: Verifier can only prepare and send quantum states to delegate a BQP computation to the prover
##[[Interactive Proofs for Quantum Computation|Quantum-authentication based verification]]: Protocols use schemes for [[Authentication of Quantum Messages|authentication of quantum messages]] to run interactive proofs.
##Trap-based based verification: Protocol uses [[Prepare-and-Send Universal Blind Quantum Computation|blind delegated quantum computation]]
##Verification based on repeated runs
#Single-prover receive-and-measure: Verifier can only receive and measure quantum states
##Measurement only verification: uses blind delegated quantum computation
##Post-hoc verification: Non-interactive (requires only single round of back and forth communication)
#Multi-prover entanglement-based: Verifier is completely classical and the provers are entangled
##Verification based on CHSH rigidity
##Verification based on self-testing graphs
##Post-hoc verification


==Properties==
==Properties==
*'''Correctness'''
[[File:Suspected relationship between BQP and MA.png|right|thumb|700px|Figure 1:Suspected relationship between BQP and MA]]
*'''Soundness'''
*'''[https://complexityzoo.uwaterloo.ca/Complexity_Zoo Complexity Classes]'''
#'''BQP''' is the class of problems which can be efficiently solved by quantum computers
#'''BPP''' is the class of problems which can be efficiently solved by classical computers.
#'''MA (Merlin-Arthur)''' is the class of problems whose solutions can be verified when given a proof setting called [[witness]].
#'''IP (interactive-proof system)''' is a generalization of MA, which involves back and forth communication between a verifier (a BPP machine) and prover (has unbounded computational power).
#Protocols 1.1, 1.2 are '''QPIP''' protocols and 2.1 is an '''MIP''' protocol. QPIP and MIP are classes of decision problems and can be decided by protocols 1.1, 1.2 and 2.1, respectively.


==Further Information==   
==Further Information==   
 
This problem remained unsolved for a completely classical client until Urmila Mahadev's protocol for classical verification of quantum protocols in 2018 (See Tags). The problem was formally addressed by Gottesman in 2004, and then promoted by Aaronson in 2007 [[Verification of Universal Quantum Computation#References||(1)]]. Emphasized by Vazirani later in 2007 from a philosophical point of view [[Verification of Universal Quantum Computation#References||(2)]], he raised the question, 'Is quantum mechanics a falsifiable theory?'
*Review Papers
*Review Papers
# [https://arxiv.org/abs/1709.06984 Gheorghiu et al (2018)]: Major portion of this functionality file has been adapted from this review 
==References==
==References==
#[https://arxiv.org/abs/1709.06984 Gheorghiu et al (2018)]
#[http://www.scottaaronson.com/blog/?p=284 The Aaronson $25.00 prize]
#[http://users.cms.caltech.edu/~schulman/Workshops/CS-Lens-2/cs-lens-2.html Vazirani (2007)]


<div style='text-align: right;'>''contributed by Shraddha Singh''</div>
<div style='text-align: right;'>''contributed by Shraddha Singh''</div>

Latest revision as of 21:18, 17 June 2019

Functionality[edit]

Quantum Computers perform task which are intractable for classical computers. The basic question here would be, "How should one verify the result of a quantum computer? This task is known as quantum verification or verification of quantum computation. BQP is the class of problems that can be solved by a quantum computer and BPP is the class of problems that can be solved by a classical computer. BPP is contained in BQP and hence, there are problems a quantum computer would solve that are intractable for a classical computer, to put it simply. Thus, if in future, an untrusted company claims to have built a quantum computer, how can the consumer be sure of the correctness of the results when he/she (the consumer) cannot compare the results predicted by the proposed quantum computer? This problem is addressed by the functionality, 'verification of quantum computers'. Verification of universal quantum computation targets every computation that can be performed by a quantum computer.

Tags: Quantum Functionality, Universal Task, Classical Verification of Universal Quantum Computation, Verification of Sub-Universal Quantum Computation, Verification of NP-complete problems

Protocols[edit]

  1. Single-prover prepare-and-send: Verifier can only prepare and send quantum states to delegate a BQP computation to the prover
    1. Quantum-authentication based verification: Protocols use schemes for authentication of quantum messages to run interactive proofs.
    2. Trap-based based verification: Protocol uses blind delegated quantum computation
    3. Verification based on repeated runs
  2. Single-prover receive-and-measure: Verifier can only receive and measure quantum states
    1. Measurement only verification: uses blind delegated quantum computation
    2. Post-hoc verification: Non-interactive (requires only single round of back and forth communication)
  3. Multi-prover entanglement-based: Verifier is completely classical and the provers are entangled
    1. Verification based on CHSH rigidity
    2. Verification based on self-testing graphs
    3. Post-hoc verification

Properties[edit]

Figure 1:Suspected relationship between BQP and MA
  1. BQP is the class of problems which can be efficiently solved by quantum computers
  2. BPP is the class of problems which can be efficiently solved by classical computers.
  3. MA (Merlin-Arthur) is the class of problems whose solutions can be verified when given a proof setting called witness.
  4. IP (interactive-proof system) is a generalization of MA, which involves back and forth communication between a verifier (a BPP machine) and prover (has unbounded computational power).
  5. Protocols 1.1, 1.2 are QPIP protocols and 2.1 is an MIP protocol. QPIP and MIP are classes of decision problems and can be decided by protocols 1.1, 1.2 and 2.1, respectively.

Further Information[edit]

This problem remained unsolved for a completely classical client until Urmila Mahadev's protocol for classical verification of quantum protocols in 2018 (See Tags). The problem was formally addressed by Gottesman in 2004, and then promoted by Aaronson in 2007 |(1). Emphasized by Vazirani later in 2007 from a philosophical point of view |(2), he raised the question, 'Is quantum mechanics a falsifiable theory?'

  • Review Papers
  1. Gheorghiu et al (2018): Major portion of this functionality file has been adapted from this review

References[edit]

  1. The Aaronson $25.00 prize
  2. Vazirani (2007)
contributed by Shraddha Singh