Sei sulla pagina 1di 25

#02 – Routing Algorithms

CCE62161 – Administrasi Jaringan


KJFD - Komputasi Berbasis Jaringan
Teknik Informatika, Fakultas Ilmu Komputer
UB Malang - 2019
Topics
• Background
• Bellman-Ford Algorithm and Distance Vector Approach
• Djikstra’s Algorithm
• Comparation Djikstra’s and Bellman-Ford Algorithm
• Spanning tree algorithm

Goal this chapter is describe how they work

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 2
Background
• Communication network is made up of nodes and links.
• Carries traffic where traffic from start node to end node
• Node :
• IP network  router
• Telephone network  end (central office)
• Optical network  optical or electro switch
• Link
• A link that connecting two nodes
• End of link called interface

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 3
Background
• An important requirement of a
communication network is to
flow or route traffic from a
source node to a destination
node
• How to do this?

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 4
Background
• An important requirement of a
communication network is to flow
or route traffic from a source node
to a destination node
• How to do this?
• Routing!
• Routing :
• user-oriented
• Network-oriented
• Additive cost property
• Non-additive cost property

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 5
Bellman–Ford algorithm And Distance Vector
• Shortest path, Centralized, and Distributed.
• Centralize View
• Node : 1, 2, 3,…. 6
• Link 1-2, 2-3, 4-5…
• d12 = Link cost between nodes 1 and 2
• D12 = Cost of the computed minimum cost path from node 1 to node 2.
• Equation

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 6
Bellman–Ford algorithm And Distance Vector

Minimum Cost!

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 7
Bellman–Ford algorithm And Distance Vector
• Distributed View
• Distance vector approach
• Equation

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 8
Dijkstra’s Algorithm
• Shortest path, Centralized, and Distributed.
• computes shortest paths to all destinations from a source
• Centralized Approach .

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 9
Dijkstra’s Algorithm

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 10
Dijkstra’s Algorithm

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 11
Dijkstra’s Algorithm
• Distribution Approach

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 12
Comparison Of The Bellman–ford Algorithm And
Dijkstra’s Algorithm
• The Bellman–Ford algorithm computes the shortest path to one
destination at a time while Dijksta’s algorithm computes the shortest
paths to all destinations (sometimes called the shortest path tree)
• ?  Home work

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 13
Path caching vs no path caching

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 14
Tree, Spanning Tree, And Steiner Tree
Algorithms
• How to solve multicasting?
• a request from one node is received by all other nodes or a subset of nodes in
the network
• Djikstra and bellman ford are for connecting between any two nodes in a
network

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 15
Tree, Spanning Tree, And Steiner Tree
Algorithms
• How to solve multicasting?
• a request from one node is received by all other nodes or a subset of nodes in
the network
• Djikstra and bellman ford are for connecting between any two nodes in a
network

• Tree!!
• A tree connects a node to other nodes (or subset of nodes) in a network but
without having any loop

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 16
Spanning Tree: Breadth First Search And Depth First
Search
• A spanning tree is a tree that connects all nodes in a network without
forming a loop (cycle)
• Two well known : BFS and DFS
• both algorithms do not take any link cost into account; in other
words, all links are treated to be of the same cost.

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 17
Spanning Tree: Breadth First Search
• Algorithm

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 18
Spanning Tree: Depth First Search
• Algorithm

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 19
Minimum Spanning Tree
• The BFS and DFS
approaches span the
network with a tree,
• but they do not
generate a minimum
spanning tree if the
link costs are other
than unit costs
• Jarník–Prim
Algorithm and
Krushkal’s algorithm.

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 20
Steiner Tree
• Steiner tree is different from a minimum spanning tree
• Considering all nodes in a network
• Consider a subset of nodes and then determine the minimum cost
tree that connects this subset of nodes
• Great for multicast communication

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 21
Steiner Tree

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 22
Steiner Tree

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 23
Summary

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 24
THE END

KJFD – Komputasi Berbasis Jaringan CCE62161 - Administrasi Jaringan #01 – Network & Routing – Page 25

Potrebbero piacerti anche