antnet: a nature inspired routing algorithm 1 lecturer: nona helmi student:

Post on 21-Jan-2016

223 Views

Category:

Documents

1 Downloads

Preview:

Click to see full reader

TRANSCRIPT

AntNet:A nature inspired routing

algorithm

1

Lecturer: Nona Helmi

Student:

Presentation Content

2

Introduction

Rapid growth of networks Increase of network communication

3EndConclusion ReferencesAntNet Algorithmintroductionintroduction Routing basics MAS-ACO

What Is Routing?

Routing is the act of moving information across a network from a source to a destination.

4EndConclusion ReferencesAntNet Algorithmintroduction Routing basicsRouting basics MAS-ACO

Tasks of Routing The tasks of a routing protocol are

determine the optimal paths route information

5EndConclusion ReferencesAntNet Algorithmintroduction Routing basicsRouting basics MAS-ACO

Design Goals

Optimality Simplicity and low overhead Rapid convergence Robustness and stability Flexibility

6EndConclusion ReferencesAntNet Algorithmintroduction Routing basicsRouting basics MAS-ACO

Agent & multi-agent system

Agent is a software that have some specification: Autonomy Reactivity Pro-activeness Social ability

7EndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACOMAS-ACO

Mobile Agent

A mobile agent is a software agent that can move between locations (mobility).

8EndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACOMAS-ACO

Nature inspired algorithms

9

PSO ABC GA ACO

EndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACOMAS-ACO

ACO introduced by Marco Dorigo (MILAN,ITALY) in his

doctoral thesis in 1992

10EndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACOMAS-ACO

M. Dorigo, Optimization, Learning and Natural Algorithms, PhD thesis, Politecnico di Milano, Italie,1992

ACO (cont.)

11

Based on ants method of finding food. Using to solve

traveling salesman routing in networks load balancing

EndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACOMAS-ACO

12

Swarm intelligence

EndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACOMAS-ACO

Stigmergy

13

A mechanism of indirect coordination between agents .

EndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACOMAS-ACO

The process that ants search the shortest path

14

(i) (ii) (iii)

EndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACOMAS-ACO

Ants & routing

15

Ants dropping different pheromones used to compute “shortest” path from source to destination(s);Advantages:

more flexible adaptation to failures and network congestion; use only local knowledge for routing and avoid costly communication of state to all network nodes.

EndConclusion ReferencesAntNet AlgorithmAntNet Algorithmintroduction Routing basics MAS-ACO

AntNet

16

AntNet: A Mobile agents Approach to Adaptive Routing. Introduced by GIANNI DI CARO and MARCO DORIGO.

G.D Caro and M.Dorigo,“ AntNet: distributed stigmergetic control for communications networks.” Journal of Artificial Intelligence Research 9 (1998), 317-365.

EndConclusion Referencesintroduction Routing basics MAS-ACO AntNet AlgorithmAntNet Algorithm

AntNet:Description Two kinds of Agents (Ant Packets)

Forward Ant. explores the network and collects

information. when reaches the destination,

changes into backward ant.Backward Ant.

goes back in the same path as forward ant.

update routing tables for all the

17EndConclusion Referencesintroduction Routing basics MAS-ACO AntNet AlgorithmAntNet Algorithm

AntNet:Description (cont.)

Two data structures stored in each network node:

Routing table (P)

An array

NjwMjbestjji ..1,),,( 2

18

mean

variancethe best trip time from i to j.

EndConclusion Referencesintroduction Routing basics MAS-ACO AntNet AlgorithmAntNet Algorithm

AntNet:Description (cont.)

19

Outgoing Links Routing table

Local traffic statistics

Network Node(i)

P11 P12 P1N

P2N

PmN,

P21 P21

Pm1 Pm2

.......

.......

Network Nodes

),,(1

211 bestw ),,(

2

222 bestw ),,( 2

NbestNN w

EndConclusion Referencesintroduction Routing basics MAS-ACO AntNet AlgorithmAntNet Algorithm

AntNet algorithm At regular interval δt,a forward agent is launched from

source toward the destination. At each node,the agent keeps the memory of the path

and the traffic condition. At each node, the next hop is selected from among all

those neighboring nodes which have not yet been visited.

20

)1|(|1

k

nndnd N

lPP

EndConclusion Referencesintroduction Routing basics MAS-ACO AntNet AlgorithmAntNet Algorithm

AntNet algorithm (cont.)

If a cycle is detected, the ant memory is popped out of stack.

When reaches destination, it generates a backward ant, transfers its stack to it and dies.

Backward ant takes the same path back using the ant stack transferred by the forward ant.

When a backward ant reaches a node k, Mk and Routing table for destination d is updated.

21EndConclusion Referencesintroduction Routing basics MAS-ACO AntNet AlgorithmAntNet Algorithm

AntNet algorithm (cont.)

Pfd = Pfd + r(1- Pfd ) Pnd = Pnd - r Pnd , n Nk, n f

22EndConclusion Referencesintroduction Routing basics MAS-ACO AntNet AlgorithmAntNet Algorithm

AntNet: Overview

23

13

4

2

5

6

Forward Ant (13)

Backward Ant (13)

At Node 2, Update Routing Information for 3

At Node 1, Update Routing Information for 3 and Update routing information for 2

EndConclusion Referencesintroduction Routing basics MAS-ACO AntNet AlgorithmAntNet Algorithm

Conclusion

24

The main characteristics of AntNet algorithm Nature inspired, Adaptivity Inherent parallelism Sclalablity

the Ant Colony Optimization can be applied to many other hard problems.

EndConclusionConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACO

References

1. B.Baran ,R.Sosa, AntNet routing algorithm for data networks based on mobile agents” Inteligencia Artificial’ (2001),75-84.

2. Dorigo M., Di Caro G.A., Gambardella L.M., "Ant Algorithms for Discrete Optimization" , Artificial Life, Vol. 5, N. 2, 1999

3. Dorigo M., Stuetzle T., Ant Colony Optimization, scholarpedia, 2010

4. Di Caro G. A. "Ant Colony Optimization and its application to adaptive routing in telecommunication networks" PhD thesis in Applied Sciences, Polytechnic School, Université Libre de Bruxelles, Brussels, Belgium, 2004

25EndConclusion ReferencesReferencesAntNet Algorithmintroduction Routing basics MAS-ACO

26

Questions, Comments?

Thank You

EndEndConclusion ReferencesAntNet Algorithmintroduction Routing basics MAS-ACO

top related