Lec 20

Lecture 20
Distance Vector and Link State Routing
Flooding
dij required complete knowledge of the network
but for actual routing algos, decisions based on local knowledge, not the complete picture of the
network
A simple local technique is flooding, in which every incoming packet is sent out on every outgoing
line except the one it arrived on
here even if some routers are damaaged this will ensure the message will reach the
destination
what is the cost though ?
each packet has a hop counter, at each transmission hop_counter-- and once it reaches zero its
killed
this prevents the packet from living forever inside the network
Not practical for sending most packets (cost), but flooding is tremendously robust
Distance Vector Routing

this computes without having complete network knowledge
for each node there is a routing table consiting of rows
each row would have 3 entries - destination , best_neighbour and cost_involved (bandwidth,
laency etc)
each router begins with information about neighbours, then table is exchanged with each other
whenever there is an update in your table you share it
eventually the updates will stop
this is an asynchronous algorythm
Also known as Bellman Ford Routing algo
Working
cost = delay
once every T sec
each router
sends to all its neighbours
a list of its estimated delays/costs to all destinations
This router also receives a similar table from its neighbours
Imagine that one of these tables has just come in from neighbor X, with Xi being X’s estimate of
how long it takes to get to router i.
updation (from slides): If the router knows that the delay to X is m msec, it also knows that it can
reach router i via X in Xi + m msec. By performing this calculation for each neighbor, a router can
find out which estimate seems the best and use that estimate and the corresponding link in its
new routing table.
Example
You know the cost to go to youur immediate neighbours

once a table is computed in one round, in the next round that table is transmitted to its nighbours
and discarded, in this round the routing table at the node is recomputed from its neighbours
Count to Infinity Problem

it always converges, but the rate is a problem
it reacts rapidly to good news, but slowly to bad news.
took too long to converge after the network topology changed
thats why it is not used now
in the leftside
it takes 4 excahanges to find best route from A to E
in the right side
initially after 4 excahnges you have the correct path
now A to B is broken
now in the next iteration of B it doesnot receive anything from A, only receives from C
C to A is 2 , so B computes it as 1+2 =3
in the next iteration C gets B -> A and D-> A as 3 hence C -> A is 4
again B recomputes it and then again C recomputes it and so on.
Basically now the table is not converging (or slow convergence)
It basically has to be all nodes -> infinity so it takes a long time
Link State Routing

DVR was used in the ARPANET, now discontinued
now replaced by link state routing
Variants of link state routing called IS-IS and OSPF are the routing algorithms that are most
widely used inside large networks and the Internet today
Working
when a router comes online, it discovers its neighbour and network addresses
set metrics to each of its neighbours
constructs a packet with this info and tells all routers in this network about what it learned
floods the network with this info
not blind flodding, but sequence number maintains if the packet has come before, if router
gets a packet from the same router with seq num < expected num (cur seq num), then that
packet is dropped
now once info is echanged, all the routers get all the info including the new one. Each router has a
global idea
now all of the routers can compute the shortest path without ambiguity
From slides
When a router is booted, its first task is to learn who its neighbors are. It accomplishes this goal
by sending a special HELLO packet on each point-to-point line.
The router on the other end is expected to send back a reply giving its name (ip address). These
names must be globally unique
The link state routing algorithm requires each link to have a distance or cost metric for finding
shortest paths
The cost to reach neighbors can be set automatically, or configured by the network operator
A common choice is to make the cost inversely proportional to the bandwidth of the link
higher bindwdth it can handle many things hence delayish :: cost is low
Link State Routing
https://www.youtube.com/watch?v=kW6zV-040SY
link : edge
state: up or down
step 1 :
sent hello packet to all neighbours
formulate the table <neighbour_node, distance>
all the nodes does this
step 2:
flooding
each node sends the table info to all other nodes
step 3:
each node gets a global knowledge i.e. get a complete topography idea
this is absent in DVR
step 4:
now to go from node i to j
apply dijkristra
single source shortest path

Lec 20

Uploaded by

Copyright:

Available Formats

Lec 20

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lec 20

Uploaded by

Copyright:

Available Formats

Lecture 20

Distance Vector and Link State Routing

Distance Vector Routing

You know the cost to go to youur immediate neighbours

Count to Infinity Problem

Link State Routing

Link State Routing

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.