Download Adaptive routing

Overview
Last Lecture
» Packet switching in Wide Area Networks
» Source: chapter 10
This Lecture
» Routing in WAN
» Source: chapter 10
Next Lecture
» X.25
» Source: chapter 10
TELE202 Lecture 6 Routing in WAN
1
Lecturer Dr Z. Huang
Routing
Complex, crucial aspect of packet
switched networks
Characteristics required
»
»
»
»
»
»
»
Correctness
Simplicity
Robustness
Stability
Fairness
Optimality
Efficiency
Costs is the major factor in routing
»
»
»
»
Length of path
Number of hops
Transport time
In-route delays
TELE202 Lecture 6 Routing in WAN
2
Lecturer Dr Z. Huang
Routing decision
Decision time
» When forwarding each packet
» When set up a virtual circuit
Decision place
» Distributed: made by each node/router
» Centralized: made by a central location
» Source: made by the sender
Routing decisions usually based on
knowledge of network (not always)
Distributed routing
» Nodes use local knowledge
» May collect info from adjacent nodes
» May collect info from all nodes on a potential
route
Central routing
» A central station collects info from all nodes
Update timing
» When is network info held by nodes updated
» Fixed - never updated
» Adaptive - regular updates
Routing strategies
» Fixed, flooding, random, adaptive
TELE202 Lecture 6 Routing in WAN
3
Lecturer Dr Z. Huang
Fixed routing
Single permanent route for each
source to destination pair
Determine routes using a least cost
algorithm
Route fixed, at least until a change in
network topology
TELE202 Lecture 6 Routing in WAN
4
Lecturer Dr Z. Huang
Flooding
Scheme
» No network info required
» Packet sent by node to every neighbor
» Incoming packets retransmitted on every link
except incoming link
» Eventually a number of copies will arrive at
destination
TELE202 Lecture 6 Routing in WAN
5
Lecturer Dr Z. Huang
Flooding
Optimizations
» Each packet is uniquely numbered so
duplicates can be discarded
» Nodes can remember packets already
forwarded to keep network load in bounds
» Can include a hop count in packets
Properties
» All possible routes are tried
– Very robust
» At least one packet will have taken minimum
hop count route
– Can be used to set up virtual circuit
» All nodes are visited
– Useful to distribute information (e.g.
routing)
Random routing
» Node selects one outgoing path for
retransmission of incoming packet
» Selection can be random or round robin
» Can select outgoing path based on probability
calculation
» No network info needed
» Route is typically not least cost nor minimum
hop
TELE202 Lecture 6 Routing in WAN
6
Lecturer Dr Z. Huang
Adaptive routing
Used by almost all packet switching
networks
Routing decisions change as
conditions on the network change
» Failure
» Congestion
Pros
» Improved performance
» Aid congestion control
Cons
» Requires info about network
» Decisions more complex, may not realize
theoretical benefits
» Tradeoff between quality of network info and
overhead
» Reacting too quickly can cause oscillation
» Too slowly to be relevant
Classification based on info sources
» Local (isolated)
» Adjacent nodes
» All nodes
TELE202 Lecture 6 Routing in WAN
7
Lecturer Dr Z. Huang
Isolated adaptive routing
Route to outgoing link with shortest
queue
Can include bias for each destination
» Q+B is used to decide the route
Rarely used - do not make use of
easily available info
TELE202 Lecture 6 Routing in WAN
8
Lecturer Dr Z. Huang
Adaptive routing - adjacent nodes
Distance Vector Routing
Developed by Bellman-Ford, also
called Bellman-Ford algorithm or
backward search algorithm
Used in ARPANET until 1979
Belongs to distributed adaptive
algorithm
Scheme
» Initially each router uses ECHO packets to
learn the distance to each of its neighbours.
» Periodically each router sends each
neighbour a list of estimated costs to each
destination
» Each router updates its tables with the best
cost to each destination according to the
information received from its neighbours
Problems
» Delay of link information
» Count-to-infinity problem (Shay’ book,pages
464-465)
TELE202 Lecture 6 Routing in WAN
9
Lecturer Dr Z. Huang
Adaptive routing - all nodes
Link State Routing
Developed to replace Distance Vector
in ARPANET
» Variants of Link State routing is now widely
used
» Similar to Distance Vector routing
It differs from Distance Vector in
» what information it exchanges and how to
store and process it.
» Link information is more timely propagated
Scheme
» A node gathers information on the status of
each link to each neighbour, e.g. bit rate,
delay time reliability, number of queued
packets
» The node builds a link state packet for each
link, containing the above information
» A node receiving link state packets forwards
them to all its neighbours except the one from
which it received the packet
» As link state packets are exchanged among
nodes, each node eventually learns about the
network topology and the cost and status of
links between network nodes
» Consequently each node can execute a
cheapest route algorithm such as Dijkstra’s
algorithm to determine its routing table
TELE202 Lecture 6 Routing in WAN
10
Lecturer Dr Z. Huang
Routing in ARPANET
First Generation
»
»
»
»
»
»
»
»
1969
Distributed adaptive
Estimated delay as performance criterion
Bellman-Ford algorithm (appendix 10a)
Node exchanges delay vector with neighbors
Update routing table based on incoming info
Doesn't consider line speed, just queue length
Queue length not a good measurement of
delay
» Responds slowly to congestion
Second Generation
»
»
»
»
»
»
1979, link state routing
Uses delay as performance criterion
Delay measured directly
Uses Dijkstra’s algorithm (appendix 10a)
Good under light and medium loads
Under heavy loads, little correlation between
reported delays and those experienced
Third Generation
»
»
»
»
1987
Link cost calculations changed
Measure average delay over last 10 seconds
Normalize based on current value and
previous results
TELE202 Lecture 6 Routing in WAN
11
Lecturer Dr Z. Huang
Hierarchical Routing
Previous algorithms require to keep
routing information of each node
» There may be too many nodes for each to
have complete routing tables.
» Consider a postal address: country, city, street,
number
Scheme used by Internet
» Nodes are divided into groups called domains,
which can be considered to be a separate and
independent network
» Routes between two nodes in a common
domain are determined using the domain's or
network's protocols
» Each domain has one or more specially
designated nodes called routers which
determine routes between domains.
» Domains may consist of subdomains each
with its own router
TELE202 Lecture 6 Routing in WAN
12
Lecturer Dr Z. Huang
Routing for Mobile Hosts
Problem: to route a packet to a mobile
host
Two kinds of users
» Migratory users: stationary users who move from
one fixed site to another from time to time but use
the network only when they are physically
connected to it
» Roaming users: compute on the run and want to
maintain their connections as they move around.
First step: how to find the mobile host?
» The world is divided into small areas
» Each area has a home agent, which keeps track of
users whose home is in the area
» Each area has one or more foreign agents, which
keep track of all mobile users visiting the area
» When a new user enters an area, his computer must
register itself with the foreign agent there, so that
the foreign agent can inform its home agent where
it is
TELE202 Lecture 6 Routing in WAN
13
Lecturer Dr Z. Huang
Routing for Mobile Hosts (cont.)
Second step: how to send a packet to
a mobile host?
» When a packet is sent to a mobile host, it is
routed to its home agent in the first place
» Because the home agent knows the address of
the host’s foreign agent, it forwards the
packet to the foreign agent which then passes
the packet to the mobile host
» Then the host’s home agent tells the sender to
henceforth send packets to the host’s foreign
agent, instead of sending to the home agent
» Then subsequent packets can now be routed
directly to the mobile host via the foreign
agent, by passing the home location entirely
TELE202 Lecture 6 Routing in WAN
14
Lecturer Dr Z. Huang
Summary
Issues in routing
» Cost
» Decision time and place
Routing classifications
»
»
»
»
Distributed
Central
Fixed
Adaptive
Routing algorithms
»
»
»
»
Fixed
Flooding
Random
Adaptive
– Isolated
– Distance Vector Routing
– Link State Routing
» Hierarchical
» Mobile
TELE202 Lecture 6 Routing in WAN
15
Lecturer Dr Z. Huang