A.

Younes

An Ant Algorithm for Solving QoS Multicast Routing Problem

A. Younes

[email protected]

Computer Science Department,

Faculty of Science, Sohag University,
Sohag, Egypt

Abstract
Many applications require send information from a source to multiple destinations through a
communication network. To support these applications, it is necessary to determine a multicast
tree of minimal cost to connect the source node to the destination nodes subject to delay
constraints. Based on the Ant System algorithm, we present an ant algorithm to find the multicast
tree that minimizes the total cost. In the proposed algorithm, the k shortest paths from the source
node to the destination nodes are used for genotype representation. By comparing the results
The expermintal results show that the algorithm can find optimal solution quickly and has a good
scalability.
Keywords: Multimedia Communication; Multicast Routing; Multicast Tree; Ant Colony
Algorithms; Bandwidth, Delay and Cost.

1. INTRODUCTION
The QoS multicast routing (QMR) problem concerns the search of optimal routing trees in the
distributed network, where messages or information are sent from the source node to all
destination nodes, while meeting all QoS requirements. This problem is NP completes [1]. Over
the past decades, many unconstrained or simple constrained multicast routing algorithms have
been developed. Typical approaches include (1) applying Dijkstra algorithm to find the shortest
path, (2) seeking the minimum network cost using Steiner tree routing algorithm, and (3) finding
multicast trees that the paths between source node and the destination nodes are connected and
their cost is minimized. A state of the art review and analysis can be found, see [1]- [3].
There are many studies that apply genetic algorithms (GA) and ant algorithms to solve the QMR
problems (with different types of QoS constraints) are increasing. In [4] [7], a heuristic GA is
used to solve the QMR problems. The algorithm acquires the solution by representing a multicast
tree as a chromosome so as to save the coding spaces and reduce the decoding operations
(compared with the binary coding mechanism). However, these approaches cannot be expanded.
If one or more nodes are added into the network, the system needs to scan all nodes again to
acquire the optimum solution. That is, previous network information cannot be transferred to the
expanded network.
A number of efficient heuristic algorithms given in [7] [9], consider a number of rigid QoS
criteria, such as bandwidth, delay, delay constraint, and packet loss rate. Chu [10], presented a
model that treats these constraints separately, add more constraints such as delay jitter and
packet loss rate, and take network expansion into account.
In [11], an efficient algorithm based on ant system is used for generating a low-cost multicast tree
subject to delay constraints. The algorithm starts with a backup-paths-set from the source node to
each destination nodes using Dijkstra Kth shortest path algorithm. Then transform the formed
procedure of the multicast tree to the Graph, and use AS to the QoS problems: when a ant move

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

156

A. Younes

from the node i to the node j depend on the corresponding probabilities function, and update the
Pheromone on Graph when every iteration finished.
In the last years, the genetic algorithms (GA) are gaining an increasing interest for solving
complex problems in the networking field, as network design [13] and unicast routing [14]. GA for
multicast routing without constraints was presented by [15] and [16], while authors in [17] and [18]
addressed the constrained problem taking into account the QoS level provided for real-time
applications in single multicast sessions. Luca and Lugi [19], presented an approach for group
multicast routing by genetic algorithm. Chen [12], proposed a new multicast routing optimization
algorithm based on Genetic Algorithms, which find the low-cost multicasting tree with bandwidth
and delay constraints.
In this paper, we propose an efficient algorithm based on ant system for generating a low-cost
multicast tree subject to delay constraints. The proposed algorithm uses a genetic algorithm given
in [4], to find the kth shortest paths from the source node to each destination nodes. Then we use
the Ant System to solve the QoS problems: when an Ant moves through a shortest path it
depends on the corresponding probabilities function and update Pheromone on that path after
finishing each iteration. The experimental results show the comparison between the proposed ant
algorithm and the genetic algorithm, [12]. Simulation results show our algorithm has features of
well performance of cost, fast convergence and stable delay.
The rest of the paper is organized as follows: Section 2 presents the problem description and
formulation. Sections 3 describe our Ant-System based QOS multicasting algorithm followed by
time complexity of the algorithm. Simulation results and comparison with other reported heuristics
are presented in Section4. Section 5 concludes the paper.

2. PROBLEM DESCRIPTION AND FORMULATION

A network is usually represented as a weighted directed graph G=(N,E), where N denotes the set
of nodes and E denotes the set of communication links connecting the nodes. |N| and |E| denote
the number of nodes and links in the network respectively. We consider the multicast routing
problem with bandwidth and delay constraints from one source node to multi-destination nodes.
Let X = {n0 , u1 , u 2 , ......u m } N be a set of from source to destination nodes of the multicast
tree. Where n0 is source node, and U= {u1 , u2 um } denotes a set of destination nodes.
Multicast tree T= (NT , ET ), where NT N, ET E, there exists the path PT (n0 , d) from source
node n0 to each destination node d U in T. e(i,j) is a link from node i N to node j N. Three
non-negative real value functions are associated with each link e(e E): cost C(e), delay D(e),
and available bandwidth B(e). The link cost function, C(e), may be either monetary cost or any
measure of the resource utilization, which must be optimized. The link delay, D(e), is considered
to be the sum of switching, queuing, transmission, and propagation delays. The link
bandwidth, B(e), is the residual bandwidth of the physical or logical link. The link delay and
bandwidth functions, D(e) and B(e), define the criteria that must be constrained.
The cost of the path PT is defined as the sum of the cost of all links in that path and can be given
by

C ( PT ) =

C (e )

(1)

ePT

The total cost of the tree T is defined as the sum of the cost of all links in that tree and can be
given by

C (T ) =

C (e )

(2)

eET

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

157

A. Younes

The total delay of the path PT(n0,d) is simply the sum of the delay of all links along PT(n0,d):

D( PT ) =

D(e),

d U

(3)

ePT ( n0 , d )

The delay of multicast tree T is the maximum value of delay in the path from source node n0 to
each destination node d U.

D (T ) = max(

D( P ),
T

d U )

(4)

ePT ( n0 ,d )

The bandwidth of the path PT(n0,d) is defined as the minimum available residual bandwidth at any
link along the path:

B( PT ) = min( B(e), e PT )

(5)

The bandwidth of the tree T is defined as the minimum available residual bandwidth at any link
along the tree:

B(T ) = min( B(e), e ET )

(6)

Assume the minimum bandwidth constraint of multicast tree is B, and the maximum delay
constraint id is D, given a multicast demand R, then, the problem of bandwidth-delay constrained
multicast routing is to find a multicast tree T, satisfying:
1. Bandwidth constraint: B (T) = B.
2. Delay constraint: D (T) = D.
Suppose S(R) is the set, S(R) satisfies the conditions
above, then, the multicast tree T which we find is:
C (T) = min (C (Ts), Ts

S(R))

3. THE PROPOSED ANT ALGORITHM

Assuming n0 is source node, and U= {u1, u2 um} denotes a set of destination nodes, the
smallest bandwidth constraint, and by the algorithm for finding the k shortest paths in reference
[4], we can find the candidate route set from source node to each destination node i (i.e. Pi ={p1,
p2,.,pn}). The proposed ant algorithm can be performed as the following steps:
Algorithm: Ant algorithm for multicast routing
1. Initialize network nodes.
a. Define the source node s and the destination nodes U= {u1, u2 um}
b. Set NC =0 (NC is a loop counter.), and put m ants to s
2. For each destination node ui U,
3. Let Pi be the set of the shortest paths for the destination node ui ( by using [4]).
4. Assign an initial value k = 0 ; to the pheromone intensity of every pk, k=1,2,n,
5. Begin the first tour;
6. Let m ants move from s to ui on Pi equally (the ants number in each path pk is equal).
7. Compute the pheromone amount left by x ants at pk ( k ) by using the following equation:

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

158

A. Younes

ktotal =

Q
* x;
Ck

(7 )

Where Ck is the cost of the path pk and is computed by Eq. (1).

8. Update the local pheromone k ;

k = (1 )k + total
k ;

(8)

Where
(0, 1] is the evaporation rate.
9. Begin a new tour
10. Set NC=NC+1;
11. Compute the corresponding probabilities function fk for each pk as follows:

[ k ] * [ k ]
;
k n

f k = [ j ] * [ j ]
jn

0
otherwise
Where k

( 9)

1
; dk is computed by using Eq. (3), and , denote the information accumulated
dk

during the movement of ants and the different effects of factors in the path selection.
12. Compute

k by using Eq. (7)

13. Update the global pheromone

14. Repeat from step 9 until NCmax

15. Compare between the values of

by using Eq. (8)

to get the best path for the destination ui (pui).

16. End For

17. Collect the all best path (pui) to get the multicast tree.

4. EXPERIMENTAL RESULTS
In this section, we show the effectiveness of the above algorithm by applying it on two examples
and compare the results which obtained by the proposed ant algorithm with the results of which
obtained by [12].
The parameters setting in the proposed algorithm as follows: ants number m = 30,
==1, and NCmax=20 (iteration numbers).

= 0.5 ,

4.1 First Example

We consider a network with 8 nodes taken from [12]. Each link represented by a triple-group (B,
D, C), given its value randomly as shown in Fig. 1.

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

159

A. Younes

(15,3,6)
1

(13,3,4)

2
(10,1,5)
(15,2,8)

(8,3,6)

(12,3,4)

(13,2,6)
4

(10,3,6)

8
(9,3,4)

(12,2,6

(9,1,3)
(12,2,6)

6
7
(10,2,4)

FIGURE 1: Network Topology Structure

Assuming the source node n0 is node 1, destination node set U= {4, 5, 7, 8} as shown in the
above figure. By the algorithm for finding the k shortest paths in reference [4] with smallest
bandwidth constraint B=10, we can find the candidate route set from source node 1 to each
destination node, as shown in Table 1.

Destination node

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

2
3
5
2
5
5
2
2
3
3
3
2
2
5
5
5
3
3
5
2
2

The shortest paths

8
7
6
4
4
6
4
4
6
7
8
2
4
8
4
4
4
4
8
6
6
6
4
4
6
8
4

6
7
6
2
6
6
7
7
4
4
2
6
7

5
6
5
8
7
7

2
2
8
7
8

8
8

5
7

TABLE 1: The candidate route set from source node 1 to each destination node
We find the multicast tree as shown in Fig.2 with cost=41.

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

160

A. Younes

5
3

8
7

FIGURE 2: The Multicast Tree obtained by the proposed Ant Algorithm

Figure 3(a, b, c, and d)) shows the number of ants on each path of the destination 4, 5, 7, and 8
respectively.
The following figures show the best path which represents the candidate route from the source
node 1 to the destination nodes. The horizontal axis represents the tour number and the vertical
axis represents the number of ants.

FIGURE 3(a): The iteration number and the number of Ants for the destination 4.

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

161

A. Younes

FIGURE 3(b): The iteration number and the number of Ants for the destination 5.

The best path for the destination node (7)

18
16

path 1-3-4-2-8-7

Ants number

14
12

path 1-3-4-6-7

10
8

path 1-2-4-6-7
path 1-2-8-7

path 1-5-6-7

path 1-5-6-4-2-8-7

2
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 1516 17 18 19 20

Iterations

FIGURE 3(c): The iteration number and the number of Ants for the destination 7.

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

162

A. Younes

The best path for the destination node (8)

35

Ants number

30

path 1-5-6-4-2-8

25

path 1-3-4-2-8

20

path 1-3-4-6-7-8
path 1-5-6-7-8

15

path 1-2-8

10

path 1-2-4-6-7-8

5
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Iterations

FIGURE 3(d): The iteration number and the number of Ants for the destination 8.
From the above figures, if we consider the figure 3(d) as an example we note that: the number of
ants in the path 1-2-8 increases from 11 to 30 during iteration 1 to iteration 20. But the number of
ants on the other paths is decreasing to 0. This means that, the path 1-2-8 is the best candidate
route from the source node 1 to the destination node 8.
Figure 4 represents the multicast tree which obtained by the genetic algorithm, [12].

1
2

4
8

FIGURE 4: The Multicast Tree obtained by [12].

By comparing the Multicast tree obtained by the proposed Ant algorithm, given in Fig. 2 and the
other one which obtained by using genetic algorithm [12], given in Fig. 4, we noted the following:
1. The bandwidth B of the path 1->3->4->8 in the tree obtained by [12] equal to 9 according
to Eq. 5 which is not 10 as it is imposed.
2. The path 1->2 isn't true, because the node 2 does not represent the destination node.
Hence, the multicast tree obtained by [12] is not correct, based on the parameters imposed.

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

163

A. Younes

4.2 Second Example

We consider a network with 8 nodes taken from [10]. Each link represented by a triple-group (D,
B, C), given its value randomly as shown in Fig.5 and compare the results with [10].
1
(18,100,9)

(9,60,2)

(3,80,3)

(7,90,6)

(3,90,3)

(12,120,4)
(8,110,21)

(10,75,8)
(6,70,4)
4

6
(9,120,1)

(9,40,7)
8

(4,13,10)

7
(12,80,12)

FIGURE 5: Network Topology Structure

Assuming the source node n0 is node 1, destination node set U= {2, 4, 5, 7} as shown in the
above figure. By the algorithm for finding the k shortest paths in reference [4] with smallest
bandwidth constraint B=70, we can find the candidate route set from source node 1 to each
destination node, as shown in Table 2.

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

164

A. Younes

Destination node

The shortest paths

4
2
3
2
3 4 2
5 6 4 2
2
6 7
8 2
6 4
2
6 4
3 2
5 6
7 8
3 4
3
2
4
5
6
4
6
4
2
4
4

1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1

3
5
5
3
3
5
5
5
3
5
5
3
5
3
3
5

1
1
1
1
1
1

3
3
5
3
3
5

5
4
6
4
2
3

7
6
7
2
8
4

5
7
8
7
6

7
7

TABLE 2: The candidate route set from source node 1 to each destination node
We find the multicast tree as shown in Fig.6 with cost=63.
1

5
3

6
4

FIGURE 6: The Multicast Tree obtained by the proposed Ant Algorithm

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

165

A. Younes

The multicast tree obtained by [10] is shown in Fig. 7.

1
2

5
3

6
4

FIGURE 7: The Multicast Tree obtained by [10]

By comparing the previous results we observe that the multicast tree obtained by the proposed
Ant algorithm is quite similar to the multicast tree obtained by [10].This means that the proposed
Ant algorithm is working properly.

5. CONCLUSION
This paper presented an Ant algorithm for solving QoS multicast routing problem based on
bandwidth and delay constraints. The proposed algorithm uses the kth shortest paths algorithm
[4], to construct the route set. Then, we have a set of paths for each destination nodes; the Ants
moving through the paths depending on the corresponding probabilities function and update the
Pheromone on the paths after finishing each iteration. Simulation results show that the proposed
algorithm has features of well performance of cost, fast convergence and stable delay. The
algorithm can guarantee the requirement of multimedia group communication for quality of
service.

6. REFERENCES
1. L. H. Sahasrabuddhe, and B. Mukherjee, Multicast Routing Algorithms and Protocols: A
Tutorial, IEEE Network, pp. 90-102, January, February 2000.
2. B. Wang, and J. C. Hou, Multicast Routing and Its QoS Extension: Problems, Algorithms,
and Protocols, IEEE Network, pp. 22-36, January/February 2000.
3. H. F. Salama, D. S. Reeves, and Y. Viniotis, Evaluation of Multicast Routing Algorithms for
Real-Time Communication on High-Speed Networks, IEEE Journal on Selected Areas in
Communications, Vol. 15(3): 332-345, 1997.
4. A. Younes, A Genetic Algorithm for Finding the K Shortest Paths in a Network, Egyptian
Informatics Journal, Vol.10(2), 2010.
5. Z. Wang, and B. Shi, Solution to QoS Multicast Routing Problem Based on Heuristic Genetic
Algorithm, Journal of Computer, 2001, Vol. 24(1): 55-61, 2001.
6. X. Zhou, C. Chen, and G. Zhu, A Genetic Algorithm for Multicasting Routing Problem,
International Conference on Communication Technology, Vol. 2, pp. 1248-1253, 2000.
7. Z. Wang, and J. Crowcroft, Quality of Service for Supporting Multimedia Applications, IEEE
Journal on Selected Areas in Communications, Vol. 14(7): 1228-1234, 1996.

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

166

A. Younes

8. L. Guo and I. Matta, QDMR: An Efficient QoS Dependent Multicast Routing Algorithm,
Proceedings of the Fifth IEEE Real-Time Technology and Applications Symposium, pp. 213222, 1999.
9. W. Wu, Multicast routing Algorithm Based on Multiple Qualities of Services,
Electronic Technologies, Vol. 266(8), 2000.

Applied

10. Chao-Hsien Chu, A Heuristic Ant Algorithm for Solving QoS Multicast Routing Problem,
Proceedings of the Evolutionary Computation Vol. 2, 2002.
11. Lin Huang, Haishan Han and Jian Hou, Multicast Routing Based on the Ant System,Applied
Mathematical Sciences, Vol. 1(57): 28272838, 2007.
12. Hua Chen, Baolin Sun, Multicast Routing Optimization Algorithm with Bandwidth and
Delay Constraints Based on GA, Journal of Communication and Computer, Vol. 2(5): 2005.
13. L. Atzori and A. Raccis, Network Capacity Assignment for Multicast Services Using Genetic
Algorithms, IEEE Communications Letters, vol. 8(6): 403-405, June 2004.
14. C.W. Ahn and R.S. Ramakrishna, A genetic algorithm for shortest path routing problem and
the sizing of population, IEEE Transaction on evolutionary computation, vol. 6(6): Dec. 2002.
15. R.H. Hwang, W.Y. Do, and S.C. Yang, Multicast Routing Based on Genetic Algorithms,
Journal of information science and engineering, vo1.16, pp.885-901, 2000.
16. R. Bhattacharya, P. Venkateswaran, S.K. Sanyal, R. Nandi, , Genetic algorithm based
efficient outing scheme for multicast networks, IEEE International Conference on Personal
Wireless Communications (ICPWC 2005), pp. 500 - 504, Jan 2005
17. L. Chen, Z. Yang, Z. Xu, A degree-delay-constrained genetic algorithm for multicast routing
tree, The Fourth International Conference on Computer and Information Technology (CIT),
pp. 1033 -1038, September 2004.
18. M. Hamdan and M.E. El-Hawary, Multicast routing with delay and delay variation constraints
using genetic algorithm, Canadian Conference on Electrical and Computer Engineering, vol.
4, pp. 2363 - 2366, May 2004.
19. Luca Sanna Randaccio and Luigi Atzori, A Genetic Algorithm Based Approach for Group
Multicast Routing, Journal of Networks, Vol. 1(4): , 2006.

International Journal of Computer Science and Security (IJCSS), Volume (5) : Issue (1) : 2011

167