quantum error correcting codes discrete math meets physics
play

Quantum Error-Correcting Codes: Discrete Math meets Physics Markus - PowerPoint PPT Presentation

Quantum Error-Correcting Codes LAWCI Latin American Week on Coding and Information UniCamp Campinas, Brazil 2018, July 2027 Quantum Error-Correcting Codes: Discrete Math meets Physics Markus Grassl Markus.Grassl@mpl.mpg.de


  1. Quantum Error-Correcting Codes LAWCI Latin American Week on Coding and Information UniCamp – Campinas, Brazil 2018, July 20–27 Quantum Error-Correcting Codes: Discrete Math meets Physics Markus Grassl Markus.Grassl@mpl.mpg.de www.codetables.de Markus Grassl – 1– 24.07.2018

  2. Quantum Error-Correcting Codes LAWCI Simple Quantum Error-Correcting Code Repetition code: | 0 � �→ | 000 � , | 1 � �→ | 111 � Encoding of one qubit: α | 0 � + β | 1 � �→ α | 000 � + β | 111 � . This defines a two-dimensional subspace H C ≤ H 2 ⊗ H 2 ⊗ H 2 bit-flip quantum state subspace no error α | 000 � + β | 111 � ( 1 ⊗ 1 ⊗ 1 ) H C 1 st position α | 100 � + β | 011 � ( X ⊗ 1 ⊗ 1 ) H C 2 nd position α | 010 � + β | 101 � ( 1 ⊗ X ⊗ 1 ) H C 3 rd position α | 001 � + β | 110 � ( 1 ⊗ 1 ⊗ X ) H C Hence we have an orthogonal decomposition of H 2 ⊗ H 2 ⊗ H 2 Markus Grassl – 27– 24.07.2018

  3. Quantum Error-Correcting Codes LAWCI Simple Quantum Error-Correcting Code s s s | ψ � | c 1 � ❣ s | 0 � | c 2 � ❣ s s | 0 � | c 3 � ☛✟ ✂ ✂ ✍ s 1 | 0 � ❣ ❣ q ☛✟ ❆ ❑ s 2 ❣ ❣ | 0 � q encoding syndrome computation measurement Error X ⊗ I ⊗ I gives syndrome s 1 s 2 = 10 Error I ⊗ X ⊗ I gives syndrome s 1 s 2 = 01 Error I ⊗ I ⊗ X gives syndrome s 1 s 2 = 11 Markus Grassl – 30– 24.07.2018

  4. Quantum Error-Correcting Codes LAWCI Linearity of Syndrome Computation Different Errors: Error X ⊗ I ⊗ I syndrome 10 Error I ⊗ X ⊗ I syndrome 01 Suppose the (non-unitary) error is of the form E = α X ⊗ I ⊗ I + β I ⊗ X ⊗ I. Then syndrome computation yields α ( X ⊗ I ⊗ I | ψ enc � ⊗ | 10 � ) + β ( I ⊗ X ⊗ I | ψ enc � ⊗ | 01 � ) . (error correction) �→ | ψ enc � ( α | 10 � + β | 01 � ) Markus Grassl – 33– 24.07.2018

  5. Quantum Error-Correcting Codes LAWCI Shor’s Nine-Qubit Code Hadamard basis: 1 1 | + � = 2 ( | 0 � + | 1 � ) , |−� = 2 ( | 0 � − | 1 � ) √ √ Bit-flip code: | 0 � �→ | 000 � , | 1 � �→ | 111 � . Phase-flip code: | 0 � �→ | + ++ � , | 1 � �→ | − −−� . Concatenation with bit-flip code gives: 1 | 0 � �→ 2 3 ( | 000 � + | 111 � )( | 000 � + | 111 � )( | 000 � + | 111 � ) √ 1 | 1 � �→ 2 3 ( | 000 � − | 111 � )( | 000 � − | 111 � )( | 000 � − | 111 � ) √ Claim: This code can correct one error, i. e., it is an [ [ n, k, d ] ] 2 = [ [9 , 1 , 3] ] 2 . Markus Grassl – 34– 24.07.2018

  6. Quantum Error-Correcting Codes LAWCI Shor’s Nine-Qubit Code Bit-flip code: | 0 � �→ | 000 � , | 1 � �→ | 111 � Effect of single-qubit errors: • X -errors change the basis states, but can be corrected • Z -errors at any of the three positions:   Z | 000 � = | 000 �  “encoded” Z -operator Z | 111 � = −| 111 � = ⇒ can be corrected by the second level of encoding Markus Grassl – 35– 24.07.2018

  7. Quantum Error-Correcting Codes LAWCI Knill-Laflamme Conditions Revisited • code with ONB {| c i �} , channel with Kraus operators { E k } • conditions: � c i | E † (i) k E ℓ | c j � = 0 for i � = j � c i | E † k E ℓ | c i � = � c j | E † (ii) k E ℓ | c j � = α kl • interpretation: (i) orthogonal states remain orthogonal under errors (ii) errors “rotate” all basis states the same way ✻ ✻ E ℓ | c j � ✗ ✄ ✄ errors | c j � ✄ = ⇒ E k | c j � ❇ ▼ α ❇ ✄ ✏ ✶ ✏✏ E k | c i � ✲ ❇ ✄ ✲ ❳❳❳❳ α ❳ ③ | c i � E ℓ | c i � Markus Grassl – 36– 24.07.2018

  8. Quantum Error-Correcting Codes LAWCI General Decoding Algorithm E 1 C E 2 C · · · E k C · · · V 0 E 1 | c 0 � E 2 | c 0 � · · · E k | c 0 � · · · V 1 E 1 | c 1 � E 2 | c 1 � · · · E k | c 1 � · · · . . . . ... . . . . . . . . V i E 1 | c i � E 2 | c i � · · · E k | c i � · · · . . . . ... . . . . . . . . � c i | E † k E l | c j � = δ i,j α k,l (1) Markus Grassl – 37– 24.07.2018

  9. Quantum Error-Correcting Codes LAWCI General Decoding Algorithm E 1 C E 2 C · · · E k C · · · V 0 E 1 | c 0 � E 2 | c 0 � · · · E k | c 0 � · · · V 1 E 1 | c 1 � E 2 | c 1 � · · · E k | c 1 � · · · rows are orthogonal as . . . . ... . . . . . . . . � c i | E † k E l | c j � = 0 for V i E 1 | c i � E 2 | c i � · · · E k | c i � · · · i � = j . . . . ... . . . . . . . . � c i | E † k E l | c j � = δ i,j α k,l (1) Markus Grassl – 38– 24.07.2018

  10. Quantum Error-Correcting Codes LAWCI General Decoding Algorithm E 1 C E 2 C · · · E k C · · · V 0 E 1 | c 0 � E 2 | c 0 � · · · E k | c 0 � · · · V 1 E 1 | c 1 � E 2 | c 1 � · · · E k | c 1 � · · · rows are orthogonal as . . . . ... . . . . . . . . � c i | E † k E l | c j � = 0 for V i E 1 | c i � E 2 | c i � · · · E k | c i � · · · i � = j . . . . ... . . . . . . . . inner product between columns is constant as � c i | E † k E l | c i � = α k,l � c i | E † k E l | c j � = δ i,j α k,l (1) Markus Grassl – 39– 24.07.2018

  11. Quantum Error-Correcting Codes LAWCI General Decoding Algorithm E 1 C E 2 C · · · E k C · · · V 0 E 1 | c 0 � E 2 | c 0 � · · · E k | c 0 � · · · V 1 E 1 | c 1 � E 2 | c 1 � · · · E k | c 1 � · · · rows are orthogonal as . . . . ... . . . . . . . . � c i | E † k E l | c j � = 0 for V i E 1 | c i � E 2 | c i � · · · E k | c i � · · · i � = j . . . . ... . . . . . . . . inner product between columns is constant as � c i | E † k E l | c i � = α k,l = ⇒ simultaneous Gram-Schmidt orthogonalization within the spaces V i Markus Grassl – 40– 24.07.2018

  12. Quantum Error-Correcting Codes LAWCI Orthogonal Decomposition E ′ E ′ E ′ 1 C 2 C · · · k C · · · E ′ E ′ E ′ V 0 1 | c 0 � 2 | c 0 � · · · k | c 0 � · · · E ′ E ′ E ′ V 1 1 | c 1 � 2 | c 1 � · · · k | c 1 � · · · rows are mutually . . . . ... . . . . . . . . orthogonal E ′ E ′ E ′ V i 1 | c i � 2 | c i � · · · k | c i � · · · . . . . ... . . . . . . . . columns are mutually orthogonal • new error operators E ′ k are linear combinations of the E l • projection onto E ′ k C determines the error • exponentially many orthogonal spaces E ′ k C Markus Grassl – 41– 24.07.2018

  13. Quantum Error-Correcting Codes LAWCI CSS Prelims: Linear Binary Codes Generator matrix: C = [ n, k, d ] 2 is a k -dim. subspace of F n 2 ⇒ basis with k (row) vectors, G ∈ F k × n = 2 C = { i G : i ∈ F k 2 } Parity check matrix: C = [ n, k, d ] 2 is a k -dim. subspace of F n 2 ⇒ kernel of n − k homogeneous linear equations, H ∈ F ( n − k ) × n = 2 2 | v H t = 0 } C = { v : v ∈ F n Error syndrome: erroneous codeword v = c + e ⇒ error syndrome v H t = c H t + e H t = e H t = Markus Grassl – 42– 24.07.2018

  14. Quantum Error-Correcting Codes LAWCI CSS Prelims: Dual of a Classical Code Let C ≤ F n 2 be a linear code. Then � � n � C ⊥ := y ∈ F n 2 : x i y i = 0 for all x ∈ C i =1 We denote the scalar product by x · y := � n i =1 x i y i . Lemma: Let C be a linear binary code. Then   � for x ∈ C ⊥ , | C | ( − 1) x · c =  ∈ C ⊥ . 0 for x / c ∈ C Markus Grassl – 43– 24.07.2018

  15. Quantum Error-Correcting Codes LAWCI Bit-flips and Phase-flips Let C ≤ F n 2 be a linear code. Then the image of the state � 1 � | c � | C | c ∈ C under a bit-flip x ∈ F n 2 and a phase-flip z ∈ F n 2 is given by � 1 ( − 1) z · c | c + x � . � | C | c ∈ C Hadamard transform H ⊗ . . . ⊗ H maps this to � ( − 1) xz ( − 1) x · b | b + z � � | C ⊥ | b ∈ C ⊥ Markus Grassl – 44– 24.07.2018

  16. Quantum Error-Correcting Codes LAWCI Proof: � 1 Hadamard transform H ⊗ n = ( − 1) a · b | b �� a | √ 2 n a , b ∈ F n 2 � � � � � 1 1 H ⊗ n ( − 1) z · c | c + x � ( − 1) z · c + a · b | b �� a | c + x � � = � | C | 2 n | C | a , b ∈ F n c ∈ C c ∈ C 2 � � 1 ( − 1) z · c +( c + x ) · b | b � = � | C | 2 n b ∈ F n c ∈ C 2 � ( − 1) x · b � 1 ( − 1) c · ( b + z ) | b � = � | C | 2 n b ∈ F n c ∈ C 2 � ( − 1) x · ( b + z ) � 1 ( − 1) c · b | b + z � = � | C | 2 n b ∈ F n c ∈ C 2 � = ( − 1) x · z ( − 1) x · b | b + z � � | C ⊥ | b ∈ C ⊥ Markus Grassl – 45– 24.07.2018

  17. Quantum Error-Correcting Codes LAWCI CSS Codes Introduced by R. Calderbank, P. Shor, and A. Steane [Calderbank & Shor PRA, 54 , 1098–1105, 1996] [Steane, PRL 77 , 793–797, 1996] Construction: Let C 1 = [ n, k 1 , d 1 ] q and C 2 = [ n, k 2 , d 1 ] q be classical linear codes with C ⊥ 2 ≤ C 1 . Let { x 1 , . . . , x K } be representatives for the cosets C 1 /C ⊥ 2 . Define quantum states � 1 | x i + C ⊥ 2 � := � | x i + y � | C ⊥ 2 | y ∈ C ⊥ 2 Theorem: Then the vector space C spanned by these states is a quantum code with parameters [ [ n, k 1 + k 2 − n, d ] ] q where d ≥ min( d 1 , d 2 ) . Markus Grassl – 46– 24.07.2018

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend