Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
1. Introduction
One of the best known problems in extremal directed graphs is the so-called degree/diameter
problem: For given numbers d and k, construct a digraph of maximum out-degree d and diameter k, with the largest possible number of vertices nd,k . In other words, find
N (d, k) = max{n : G(n, d, k) 6= }.
A straightforward general upper bound on the maximum order nd,k of a directed graphs
(digraph) of maximum degree d and diameter k is the Moore bound Md,k for directed graphs.
nd,k Md,k = 1 + d + d2 + + dk
( k+1
d
1
if d > 1
d1
=
k+1
if d = 1
The equality nd,k = Md,k holds only when d = 1 or k = 1 [19][17], and hence in all other
cases the upper bound can be lowered by 1.
Case k = 2. It is well known that the bound Md,2 1 can be achieved for all d 2 by line
digraphs of complete digraphs of order d + 1. The existence of digraphs of Md,2 1 vertices,
other than the line digraphs, has been studied in [8][3][5].
Case k 3. In general, it is not known whether or not Md,k 1 is attainable. The
existence of digraphs of degree d, diameter k 3 and order Md,k 1 has been studied and
several necessary conditions have been given in [3][5]. For degree d = 2, it was shown in [13]
that M2,k 1 is not attainable. For degree d = 3, it was proved that digraphs with M3,k 1
nodes do not exist [7]. Moreover, it was shown in [15] that M2,k 2 cannot be attained for
k 3.
Hence in many cases (depending on k and d) the upper bound on nd,k is actually 2 or 3
less than the Moore bound. But apart from that, no other upper bounds on nd,k are known.
A general lower bound on the largest order nd,k for the degree/diameter problem is given
by Kautz digraphs K(d, k) [20] of order dk + dk1 ; these digraphs can be obtained by (k 1)fold iteration of the line digraph construction applied to the complete digraph of order d + 1.
It is also known that n2,4 = 25 (which implies that n2,j 25 2j4 for j 4 by the iterated
line digraph construction); the corresponding digraph of order 25 was found by Alegre.
In view of the huge gap between the Moore bound and the best lower bound, much effort
has been spent in generating large digraphs of given degree and diameter. In the next section
we give an overview of construction techniques for large digraphs. Finally, in the last section
we list open problems in the degree/diameter problem area for digraphs.
3
0
8
5
7
4
2
0
4
7
1
3
1
4
1
5
2
8
10
6
2
3
4
11
12
L2 (G)
L(G)
Figure 3: Digraphs G G(3, 2, 1) with its line digraphs L(G) G(6, 2, 2) and L2 (G)
G(12, 2, 3).
w N + (v) in G, where
v0 =
v,
if uw V (LG)
For example, Figure 4(a) shows the digraph G with a vertex-set V = {0, 1, 2, 3, 4, 5} and
an arc-set
A = {01, 03, 12, 15, 20, 23, 34, 35, 40, 41, 52, 54}.
Let A0 = {01, 03, 12, 15, 20, 23, 34, 35, 41, 52, 54} (shown as solid lines in Figure 4(a)) is a
subset of A. Then a partial line digraph LG of the digraph G has the vertex-set
V (LG) = {01, 03, 12, 15, 20, 23, 34, 35, 41, 52, 54}
as shown in Figure 4(b). In this example, vertices 34 and 54 are adjacent to vertex 20 because
the arc 40
/ A0 .
12
23
52
01
15
3
(a)
35
20
03
54
41
(b)
34
x02 =
x2 ,
if x2 x3 ...xk V
2
x002 x3 ...xk V ,
otherwise
012
101
120
201
020
121
210
021
010
102
Figure 5: Example of generalised Kautz digraph on alphabet.
d
e
5
6
1
b
c
3
14
11
12
15
16
18
13
10
17
Figure 8(a) shows an example of digraph G G(12, 2, 3) with the property that some
vertices have identical out-neighbourhoods. For example, N + (7) = N + (11). Deleting vertex
12 together with its outgoing arcs and then reconnecting its incoming arcs to vertex 7 (since
N + (7) = N + (11)), we obtain a new digraph G1 G(11, 2, 2) as shown in Figure 8(b).
yz
r
q
o
p
t
o
u
v
v
x
G0
0
7
10
4
3
11
(a)
9
7
10
4
2
(b)
3. Open Problems
A frequent goal in constructions of large digraphs with given properties is to keep the
diameter as small as possible. The voltage assignment technique seems a very promising
technique for finding large digraphs. Preliminary results regarding the voltage assignment
technique are extremely encouraging, see [6] and [1]. In [6] the authors presented an upper
bound on the diameter of lifted digraphs, applicable to a fairly general class of base digraphs
and general groups.
Table 1-4 summarise our knowledge of best current lower bound Nl (d, k) and upper bound
Nu (d, k) for 2 k 10 and d = 2, 3, 4 and 5. The tables also show the range of the potential
larger orders nd,k . The huge gap between the upper bound and the best lower bound strongly
represent an open problem.
Using the voltage assignment technique or some other established technique, possibly with
the aid of computers, there is a good chance that a new digraphs with larger order can be
found.
k
2
3
4
5
6
7
8
9
10
26,27,28
n2,k = Nl (2, k) + i, i = 1 . . . 10
n2,k = Nl (2, k) + i, i = 1 . . . 24
n2,k = Nl (2, k) + i, i = 1 . . . 52
n2,k = Nl (2, k) + i, i = 1 . . . 108
n2,k = Nl (2, k) + i, i = 1 . . . 220
n2,k = Nl (2, k) + i, i = 1 . . . 444
k
2
3
4
5
6
7
8
9
10
37, 38
n3,k = Nl (3, k) + i, i = 1 . . . 11
n3,k = Nl (3, k) + i, i = 1 . . . 38
n3,k = Nl (3, k) + i, i = 1 . . . 119
n3,k = Nl (3, k) + i, i = 1 . . . 362
n3,k = Nl (3, k) + i, i = 1 . . . 1, 091
n3,k = Nl (3, k) + i, i = 1 . . . 3, 278
n3,k = Nl (3, k) + i, i = 1 . . . 9, 839
k
2
3
4
5
6
7
8
9
10
k
2
3
4
5
6
7
8
9
10
References
[1] B. McKay, M. Miller and J. Siran, A note on large graphs of diameter two and given
maximum degree, J. Combinatorial Theory(B) 74 (1998) 110-118.
an
[2] E.T. Baskoro, L. Brankovic, M. Miller, J. Plesnk, J. Ryan and J. Sir
, Large digraphs
with small diameter: A voltage assignment approach, JCMCC 24 (1997) 161176.
Znam, Digraphs of degree 3 and order close
[3] E.T. Baskoro, M. Miller, J. Plesnk and S.
to Moore bound, J. Graph Theory 20 (1995) 339-349.
[4] E.T. Baskoro, M. Miller and J. Plesnk, On the structure of digraphs with order close to
the Moore bound, Graphs and Combinatorics 14 (1998) 109-119.
[5] E.T. Baskoro, M. Miller and J. Plesnk, Further results on almost Moore digraphs, Ars
Combinatoria, in press.
[6] E.T. Baskoro, L. Brankovic, M. Miller, J. Plesnik, J. Ryan and J. Siran, Large digraphs
with small diameter: A voltage assignment approach, JCMCC 24 (1997) 161-176.
[7] E.T. Baskoro, M. Miller, J. Siran and M. Sutton, A complete characterisation of Almost
Moore digraphs of degree 3, preprint, 1989.
Znam, Regular digraphs of diameter 2 and
[8] E.T. Baskoro, Mirka Miller, J. Plesnk and S.
maximum order, Australasian J. Combin. 9 (1994) 291-306. See also Errata 13 (1995).
[9] M.A. Fiol, A.S. Llado and J.L. Villar, Digraphs on alphabets and the (d, N ) digraph
problem, Ars Combinatoria 25C (1988) 105122.
[10] M.A. Fiol and A.S. Llado, The partial line digraph technique in the design of large
interconnection networks, IEEE Transactions on Computers 41 (1992) 848857.
[11] M. Imase and M. Itoh, Design to minimize diameter on building-block network, IEEE
Trans. on Computers C-30 (1981) 439442.
[12] M. Miller, Diregular digraphs with minimum diameter, Master of Arts Thesis, Dept. of
Mathematics, Statistics and Computing Science, UNE Armidale Australia (1986).
[13] M. Miller and I. Fris, Maximum order digraphs for diameter 2 or degree 2, Pullman
volume of Graphs and Matrices, Lecture Notes in Pure and Applied Mathematics 139
(1992) 269278.
[14] M. Miller and I. Fris, Minimum diameter of diregular digraphs of degree 2, Computer
Journal 31 (1988) 71-75.
[15] M. Miller and J. Siran, Digraphs of degree two and defect two, Discrete Mathematics, in
press.
[16] M. Miller and Slamin, On the monotonocity of minimum diameter with respect to order
and maximum out-degree, Proceeding of COCOON 2000, Lecture Notes in Computer
Science 1558 (D.-Z Du, P. Eades, V.Estivill-Castro, X.Lin (eds.)) (2000) 193-201.
Znam, Strongly geodetic directed graphs, Acta Fac. Rer. Nat. Univ.
[17] J. Plesnk and S.
Comen., Math. 29 (1974) 29-34.
[18] Slamin, Diregularity of Digraphs Close to Moore Bound, Ph.D. Stud. Thesis, The University of Newcastle, Australia (2001).
[19] W. G. Bridges and S. Toueg, On the impossibility of directed Moore graphs, J. Combinat.
Theory 29 (1980) 339-341.
[20] W.H. Kautz, Bounds on directed (d, k) graphs, Theory of cellular logic networks and
machines. AFCRL-68-0668 Final Report (1968) 2028.