Perfect codes in Generalized

Sierpiński Graphs
Aline Parreau
Institut Fourier - Université de Grenoble - France

Join work with Sylvain Gravier and Matjaz Kovše

CID 2011- September 20th , 2011

ANR IDEA

1/21
Outline
Sierpiński Graphs:

→ Graph on {1, ..., k}n with good metric and coding properties.

Idea : generalize those graphs to have new (and good?) metrics on

{1, ..., k}n

2/21
Recursive construction of Sierpiński graph S(n, k)

1. Start with the complete graph: S(1, k) = Kk .

1 2

3/21
Recursive construction of Sierpiński graph S(n, k)
1. Start with the complete graph: S(1, k) = Kk .
2. Copy it k times.

3/21
Recursive construction of Sierpiński graph S(n, k)
1. Start with the complete graph: S(1, k) = Kk .
2. Copy it k times.
3. Add one edge between each pair of copies to get S(2, k).

3/21
Recursive construction of Sierpiński graph S(n, k)
1. Start with the complete graph: S(1, k) = Kk .
2. Copy it k times.
3. Add one edge between each pair of copies to get S(2, k).
4. ”New” vertex i is vertex i of copy i.

1 2

3/21
Recursive construction of Sierpiński graph S(n, k)
1. Start with the complete graph: S(1, k) = Kk .
2. Copy it k times.
3. Add one edge between each pair of copies to get S(2, k).
4. ”New” vertex i is vertex i of copy i.
5. Repeat to obtain S(3, k),S(4, k),...

1 2 1 2

3/21
Examples of Sierpiński graphs

S(n, k) : k vertices in the complete graph, n iterations

11
00
00
11
11
00 1111
0000
11
00 1
0 1111
0000
1
0 1
0 1111
0000
1
0 1
0 1
0 00
11
11
00 00
11 1
0 0
1 1
0 0
1 1
0 1
0
11
00
00
11
1
0 00
11
0
1
0
1
11
00 0000
1111
11
00 1
0 0
1
1
0 0
1
1
0 0000
1111
1
0 1
0 0
1
1
0 0
1
0
1
00
11
0
1 0000
1111
00
11 1
0
0000
1111 0
1
0
1
0
1 0
1
0
1
0
1 0000
1111
0
1 1
0
0000
1111 0
1
0
1
0
1 11
00
0
1
1 0000
1111 0
1 0
1 0000
1111 0
1 00
11
00
11
0
0
1
1111
0000
0000
1111
1
0
0
1
1
0
0
1
1111
0000
0000
1111
1
0
0
1 00
11
11
00
00
11
0
1
00
11 0000
1111
00
11 0
1 0
1
0
1 0
1
0
1 0000
1111
0
1 0
1 0
1
0
1 11
00
00
11 1
0
0
1
00
11 0000
1111
00
11 0
1 0
1
0
1 0
1
0
1 0000
1111
0
1 0
1 0
1
0
1 1
0
1
0
0
1
00
11 0000
1111
00
11 0
1 0
1
0
1 0
1
0
1 0000
1111
0
1 0
1 0
1
0
1 00
11
11
00 11
00
11
00 00
11
11
00
00
11 00
11 0
1 0
1 0
1 0
1 0
1 0
1 11
00 11
00
00
11
0
1
11
00
0
1 1111
0000
11
00 0
1 0000
1111
1
0 0
1
0000
1111 1111
0000
1
0 0
1 0
1
1
0
0
1
11
00 11
00 00
11
00
11
0
1 00
11 0
1 0
1 0
1
0000
1111 0
1 0
1 0
1
0
1 0
1
1
0 00
11
00
11
0
1 0000
1111 0
1 0
1
0
1
0
1 0000
1111
0000
1111 0
1
0
1
1
0
00
11 00
11 0
1 1111
0000
0
1 0
1 0
1 0
1 1
0
0
1 11
00
0
1
00
11 1111
0000
11
00 0
1 0000
1111
0
1 0
1 1111
0000
1
0 0
1 0
1
0
1 00
11 00
11
0
1
00
11 00
11 0
1 0000
1111
0
1 0
1 0
1 0
1 0
1
0
1 11
00 00
11
00
11
00
11
11
00
00
11
11
00
0
1
00
11 11
00
00
11 0
1
0000
1111
0
1 1
0
0
1
0
1 1
0
0
1
0
1 1
0 0
1
0000
1111
0
1 0
1 1
0
0
1
0
1 00
11
0
1
00
11 0000
1111
00
11 0
1 0
1
0
1 0
1
0
1 0000
1111
0
1 0
1 0
1
0
1 11
00
0
1
0
1 0000
1111
0000
1111 0
1
0
1 0
1
0
1 0000
1111
0000
1111 0
1
0
1
00
11
11
00 00
11 00
11
0
1 0000
1111 0
1 0
1 0000
1111 0
1 00
11 11
00
00
11
0
1 0000
1111 0
1 0
1 0000
1111 0
1
0
1
00
11
0
1 0000
1111
00
11 0
1
0000
1111 0
1
0
1
0
1 0
1
0
1
0
1 0000
1111
0
1 0
1
0000
1111 0
1
0
1
0
1
00
11
11
00
00
11
0
1
00
11 00
11
00
11 0
1
0000
1111
0
1 0
1
0
1
0
1 0
1
0
1
0
1 0
1 0
1
0000
1111
0
1 0
1 0
1
0
1
0
1 00
11 00
11
00
11
11
00
00
11 00
11 11
00
11
00 00
11
00
11
00
11 00
11 0
1 0
1 0
1 0
1 0
1 0
1 11
00
11
00
11
00 1111
0000
11
00 0
1 1111
0000
1
0 0
1 1111
0000
1
0 0
1 1
0 11
00 11
00
00
11 00
11 0
1 0
1 0
1 0
1 0
1 0
1 11
00
11
00

S(4,3) S(5,2)

4/21
Sierpiński graphs: definition with words

Vertex set of Sierpiński graphs: {1, . . . , k}n

Edge between u1 u2 ...un and v1 v2 ...vn , if there is 1 ≤ j ≤ n s.t:

• ui = vi if i < j,
• uj 6= vj ,
• ui = vj and vi = uj if i > j

u= w x y ...y
v= w y x ...x

Extreme vertex x: vertex x...x

5/21
Sierpiński graphs: definition with words
333

331 332

313 323

311 322
312 321
133 233

131 132 231 232

113 123 213 223

111 222
112 121 122 211 212 221

u= w x y ...y
v= w y x ...x

6/21
About Sierpiński Graphs

• Introduced in 1997 by Klavžar and Milutinović

• S(n, 3) are Hanoı̈ graphs (n = number of disks):

7/21
About Sierpiński Graphs

• Introduced in 1997 by Klavžar and Milutinović

• S(n, 3) are Hanoı̈ graphs (n = number of disks):

7/21
About Sierpiński Graphs

• Introduced in 1997 by Klavžar and Milutinović

• S(n, 3) are Hanoı̈ graphs (n = number of disks):

7/21
About Sierpiński Graphs

• Introduced in 1997 by Klavžar and Milutinović

• S(n, 3) are Hanoı̈ graphs (n = number of disks):

• S(n, k): Hanoı̈ game on k rods and n disks with another move:

7/21
Study of Sierpiński Graphs

• Metric properties (shortest path, diameter. . . ), (Klavžar, Milutinović,

1997)
• Hamiltonicity, (Klavžar, Milutinović, 1997)
• Automorphism group, (Klavžar, Mohar, 2000)
• Crossing number, (Klavžar, Mohar, 2000)
• Perfect codes, identifying codes, (a, b)-codes,... (Klavžar et al. 2002,
Beaudou et al. 2010,. . . )
• ...

8/21
Generalization

⇒
Complete graph Kk General graph G on k vertices

S(n, k) S(n, G )

9/21
Recursive construction
5
copy 5
4 3

5 5
1 2
4 3 4 3
5
4 3
→ copy 4
1 2 1 2
copy 3

1 2
5 5
4 3 4 3
copy 1 copy 2

1 2 1 2
• A copy for each vertex of G

10/21
Recursive construction
5
copy 5
4 3

5 5
1 2
4 3 4 3
5
4 3
→ copy 4
1 2 1 2
copy 3

1 2
5 5
4 3 4 3
copy 1 copy 2

1 2 1 2
• A copy for each vertex of G
• Edge ij in G → edge between i in copy j and j in copy i

10/21
Recursive construction
5
copy 5
4 3

5 5
1 2
4 3 4 3
5
4 3
→ copy 4
1 2 1 2
copy 3

1 2
5 5
4 3 4 3
copy 1 copy 2

1 2 1 2
• A copy for each vertex of G
• Edge ij in G → edge between i in copy j and j in copy i

10/21
Recursive construction
5

5 4 3

4 3
→
1 2

1 2
• A copy for each vertex of G
• Edge ij in G → edge between i in copy j and j in copy i
• Extreme vertex ( ) : vertex i of copy i
10/21
Recursive construction
11
00
11
00
11
00 11
00
11
00 11
00

1
0 00
11 11
00 1
0
0
1 00
11 00
11 0
1
0
1 00
11 00
11 0
1
1
0 1
0 1
0 11
00
1
0 1
0 1
0 11
00

1
0
0
1 1
0
0
1 1
0
0
1 11
00
00
11

1
0 1
0
11
00 1
0 0
1 11
00
11
00 1
0 1
0 1
0 00
11 11
00
1
0 11
00 0
1 0
1 0
1 00
11 11
00 1
0
1
0 11
00 11
00 1
0
1
0 1
0 1
0 11
00
0
1 0
1 00
11 0
1 0
1
0
1 0
1
0
1 0
1
0
1 00
11
00
11 0
1 00
11 0
1 00
11
1
0
0
1 1
0
0
1 11
00
00
11 1
0
0
1 1
0
0
1 00
11
00
11 1
0
0
1 00
11
00
11
11
00 1
0 00
11 1
0 1
0 00
11 1
0 1
0
00
11
00
11 0
1
0
1 00
11
00
11 0
1
0
1 0
1
0
1 00
11
00
11 0
1
0
1 0
1
0
1

11
00
00
11 1
0
0
1 11
00
00
11 1
0
0
1 1
0
0
1 11
00
00
11 1
0
0
1 1
0
0
1

1
0
1
0 1
0
1
0 1
0
0
1 11
00
11
00
11
00 1
0 11
00 1
0 1
0 11
00 1
0 1
0
11
00 1
0 11
00 1
0 1
0 11
00 1
0 1
0

00
11 0
1 00
11 0
1 0
1 00
11 0
1 0
1
11
00 1
0 11
00 1
0 1
0 11
00 1
0 1
0
00
11 0
1 00
11 0
1 0
1 00
11 0
1 0
1

00
11 00
11
00
11 00
11
00
11 00
11
0
1 00
11 00
11 0
1
1
0
0
1 00
11
00
11 11
00
00
11 0
1
0
1

1
0 1
0 11
00 1
0 1
0 11
00 1
0 11
00
1
0 1
0 00
11 0
1 0
1 11
00 1
0 11
00
11
00
00
11 1
0
0
1 00
11
00
11 1
0
0
1 1
0
0
1 00
11
00
11 1
0
0
1 0
1
0
1

11
00
00
11 1
0
0
1 11
00
00
11 1
0
0
1 1
0
0
1 11
00
00
11 1
0
0
1 1
0
0
1
00
11 0
1 00
11 0
1 0
1 00
11 0
1 0
1

0
1 0
1 0
1 00
11
1
0 1
0 1
0 00
11
0
1 0
1 0
1 00
11
11
00 1
0 11
00 1
0 1
0 00
11 1
0 1
0
00
11 0
1 00
11 0
1 0
1 00
11 0
1 0
1
00
11 0
1 00
11 0
1 0
1 00
11 0
1 0
1

11
00 1
0 11
00 1
0 1
0 11
00 1
0 1
0
11
00 1
0 11
00 1
0 1
0 11
00 1
0 1
0

10/21
Equivalent definition with words

Vertex set: {1, . . . , k}n

Edge between u1 u2 ...un and v1 v2 ...vn , if there is 1 ≤ j ≤ n s.t:

• ui = vi if i < j,
• uj 6= vj , and (uj , vj ) ∈ E (G )
• ui = vj and vi = uj if i > j

u= w x y ...y
v= w y x ...x (x, y ) ∈ E (G )

Extreme vertex x: vertex x...x

11/21
Basic properties

Idea: express property on S(n, G ) with property of G

• Chromatic number: χ(S(n, G )) = χ(G ).

• Distance between extreme vertex i and
extreme vertex j: (2n − 1) ∗ dG (i, j).
• Connexity: a cutting edge in G implies
some cutting edges in S(n, G ).
• Self-similar graph: vertices starting with
u,|u| = i induce S(n − i, G ).

12/21
Perfect codes in classical Sierpiński graphs

Theorem (Klavžar, Milutinović, Petr,2002)

There is always a perfect code in S(n, Kk )

Each vertex is dominated by exactly one code vertex

13/21
General case

Proposition
If any packing of G leaves at least two vertices, S(n, G ) does never
have a perfect code.

Proposition
If G does not have a perfect code, there is an equivalence between:
• S(n, G ) has a perfect code for one n > 1.
• S(n, G ) has a perfect code for any n > 1.
• S(2, G ) has a perfect code.
• G has a ’special’ 2-factor.

14/21
Special case: Power of cycles

We consider G = Ckr :

k = i[2r + 1] r n=1 n=2 n>3

i >1 no no no
i =1 r even no no no
i =1 r odd no yes yes
i =0 1 < r < n2 yes yes no
n
i =0 r = 1 or r ≥ 2 yes yes yes

15/21
Elements of proof

We have k = 1[2r + 1]. Therefore:

• G = Ckr does not have a perfect code.
• There is a packing of G that let only one vertex not dominated.

We have to show that:

S(2, G ) has a perfect code ⇒ r is odd

16/21
Elements of proof
Assume there is a perfect code C in S(2, G ), consider a copy of G :

copy i

...

• C induces a packing in copy i of G ,

17/21
Elements of proof
Assume there is a perfect code C in S(2, G ), consider a copy of G :

copy i c(i)

...

• C induces a packing in copy i of G ,

• c(i) : bridge vertex in C

17/21
Elements of proof
Assume there is a perfect code C in S(2, G ), consider a copy of G :

i
a(i)

copy i c(i)

...

• C induces a packing in copy i of G ,

• c(i) : bridge vertex in C
• a(i) : bridge vertex not dominated inside copy i

17/21
Elements of proof
Assume there is a perfect code C in S(2, G ), consider a copy of G :

i
a(i)

c(a(i))
copy i c(i)

a(c(i))
...

• C induces a packing in copy i of G ,

• c(i) : bridge vertex in C
• a(i) : bridge vertex not dominated inside copy i
• |a(i) − c(i)| = r + 1, a ◦ c = c ◦ a = id

17/21
Elements of proof
• For each i : a(i),c(i),
• |a(i) − c(i)| = r + 1, a ◦ c = c ◦ a = id

i
a(i)

c(i)

18/21
Elements of proof
• For each i : a(i),c(i),
• |a(i) − c(i)| = r + 1, a ◦ c = c ◦ a = id

i
a(i)

c(i)

• Construction of a 2-factor on {1, . . . , k}.

18/21
Elements of proof
• For each i : a(i),c(i),
• |a(i) − c(i)| = r + 1, a ◦ c = c ◦ a = id

• Construction of a 2-factor on {1, . . . , k}.

• Induces a mapping on {1, . . . , r + 1}, so r + 1 is even.

18/21
When G has a perfect code

k = i[2r + 1] r n=1 n=2 n>3

i >1 no no no
i =1 r even no no no
i =1 r odd no yes yes
i =0 1 < r < n2 yes yes no
n
i =0 r = 1 or r ≥ 2 yes yes yes

We study weak perfect code: packing with only extreme vertices not
dominated.
• Caracterized by the status of extreme vertices: only few possibilities.
• Weak perfect code for n give weak perfect code for n + 1.
• no weak perfect code for some n0 ⇒ no perfect code for n ≥ n0 .

19/21
Result on the automorphism group
• Description and size:

id

id

20/21
Result on the automorphism group
• Description and size:

id

id

n−2
|Aut(n, C4 )| = O(24 +...
)

20/21
Result on the automorphism group
• Description and size:
n−2
|Aut(n, C4 )| = O(24 +...
)

• Distinguishing number
→ with the distinguishing number of G :

20/21
Result on the automorphism group
• Description and size:
n−2
|Aut(n, C4 )| = O(24 +...
)

• Distinguishing number
→ with the distinguishing number of G :
0
1 00
11 0
1 00
11
1
0
0
1 11
00
00
11 1
0
0
1 11
00
00
11

1
0 11
00 1
0 11
00
0
1 00
11 0
1 00
11

1
0 11
00 1
0 11
00
0
1
0
1 00
11
00
11 0
1
0
1 00
11
00
11

1
0 11
00 1
0 11
00
0
1 00
11 0
1 00
11

D(S(n, G )) = max(max D(G /x), 2)

x∈V

20/21
Perspectives

• Hamiltonicity,
• Crossing number,
• Identifying codes, (a, b)-codes,
• Any local property,...

21/21
Perspectives

• Hamiltonicity,
• Crossing number,
• Identifying codes, (a, b)-codes,
• Any local property,...

Thank you for your attention !

21/21