1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational...

9
1 Quantum query complexity of some graph problems C. Dürr Univ. Paris-Sud M. Heiligman National Security Agency P. Høyer Univ. of Calgary M. Mhalla Institut IMAG, Grenoble

Transcript of 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational...

Page 1: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

Quantum query complexity of some graph problems

C. Dürr Univ. Paris-SudM. Heiligman National Security AgencyP. Høyer Univ. of Calgary M. Mhalla Institut IMAG, Grenoble

Page 2: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

Single Source Shortest Paths

• •• •

• •

••

• •

••

••

Given a directed graph G(V,E), with non-negative edge weights and a source vertex v0

find the shortest paths to all vertices v

v01 1 4 3 12

How many queries of the type ''what is the weight of the edge (u,v)?'' are necessary to solve the problem with bounded error?

Classical (n2)Quantum(n3/2), O(n3/2log3/2n)

Page 3: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

Single Source Shortest Paths

• •• •

• •

••

• •

••

••

Given a directed graph G(V,E), with non-negative edge weights and a source vertex v0

find the shortest paths to all vertices v

v01 1 4 3 12

How many queries of the type ''what is the weight of the edge (u,v)?'' are necessary to solve the problem with bounded error?

Classical (n2)Quantum(n3/2), O(n3/2log3/2n)

Page 4: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

General algorithm

• •• •

• •

••

• •

••

••

Tree T={v0} covering vertices S={v0}while |S|<n

add cheapest border edge (u,v)∈E∩Sx(V\S) to A

add v to S

Definition cost of edge (u,v) =shortest path weight(v0,u) + edge

weight(u,v)

v0

Page 5: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

P3

Quantum procedure for finding cheapest border edgeConsider the decomposition of |S| into powers of 2Decompose S into P1∪…∪Pk s.t.

●|P1|>…>|Pk| ●and each |Pi| is a power of 2

P1 P2• •

• •

• •

••

••

••

Page 6: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

P3

Quantum procedure for finding cheapest border edgeConsider the decomposition of |S| into powers of 2Decompose S into P1∪…∪Pk s.t.

●|P1|>…>|Pk| ●and each |Pi| is a power of 2●Suppose for every Pi we computed Ai : the |Pi| cheapest border edges of Pi with distinct targets(for edges with source∈Pi and target∉P1∪…∪Pi)

P1 P2• •

• •

• •

••

A1 A2

A3

••

••

Page 7: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

P3

Observations●Ai∩Sx(V\S) (restricted to targets∉S) is non empty for every i ●The cheapest border edge of S (u,v) has its source u∈Pi for some i, and therefore v∈Ai●Thus (A1∪…∪Ak)∩Sx(V\S) contains the cheapest border edge of S

P1 P2• •

• •

• •

••

A1 A2

A3

uv

••

••

Page 8: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

Computing Ak usinga minimum search procedure

P3

P1 P2• •

• •

• •

••

• •

••

•• 5• 2

8

9

Inputmatrix ℕa×b

Outputa column disjoint minimal entriesBounded error quantum query complexity(ab)

8 5 ∞∞

∞ 2 9∞

Page 9: 1 Quantum query complexity of some graph problems C. DürrUniv. Paris-Sud M. HeiligmanNational Security Agency P. HøyerUniv. of Calgary M. MhallaInstitut.

1

Single source (n2) (n3/2), O(n3/2log2n) (m) ((nm)), O((nm)log2n)shortest pathsMinimum weight (n2) (n3/2) (m) ((nm))spanning treeConnectivity (n2) (n3/2) (m) (n)(undirected graph)Strong Connectivity (n2) (n3/2)(m) ((nm)), O((nmlogn))(directed graph)

Bounded error quantum query complexity

Adjacency matrix model

1: 2: 3: 4:1: 0 1 1 02: 1 0 0 03: 1 1 0 14: 0 0 0 0

Bounded error (classical) quantum query complexity

1

32

4

Adjacency array model1: 2 32: 13: 1 4 24: