Ns Simulation of IEEE 802.

11

Er. Niranjan Singh

 Project descriptions
 Goals
 Understand the IEEE 802.11
 Do wireless LAN simulations using Ns

 Focus
 Ad hoc networking
 Collision avoidance (RTS/CTS handshake)

2
 Wired vs. Wireless
 Wireless communication
 No wired links: radio, infrared, laser
 Ad hoc network
 Problems in Wireless Network (IEEE 802.11)
 No multi-hop awareness
 Hidden/ Exposed
 Unfairness
 Packet drop is occurred often by errors in transmission
layer
 (Compare) Problems in Wired network
 Major cause of dropped packets: Congestion in Routers

3
 Hidden/ Exposed node

From “The deaf node problem in Ad hoc wireless LANs”

by S. Ray, D. Starobinski, and J.B.Carrunthers

 Data transmission from A to B

 Hidden node =D (possibly Deaf node)
 Cause packet collision
 Exposed node=C
 Prohibited from transmitting

4
 802.11 Operations (#1)
1 2
B B
RTS

S R RTS
S R
A A
RTS
C C

3 4
CTS
B B
S R CTS
S R
A CTS
A
C C

5
 802.11 Operation (#2)
RTS B CTS
Data
RTS
RTS
CTS CTS
A S R C
Data Data
ACK

 Receive RTS: Defer until CTS should have been sent

 Receive CTS: Defer until Data should have been sent
 If you don’t receive CTS or ACK, back off and try it all over again

(from http://www-ece.rice.edu/!ashu/reneclass/lectures/elec437lecture2.pdf)

6
 Ns (Network Simulator)
 A discrete event simulator targeted at
networking research
 The collaboration of USC/ISI, LBL, UCB, and
Xerox PARC
 Two main components: Ns, Nam
 Validation is needed

7
 Ns
 Support wired/wireless models
 Traffic models and applications
 Web, FTP, telnet, constant-bit rate, stochastic
 Transport protocols
 Unicast: TCP(Reno, Vegas, etc.), UDP
 Multicast: SRM
 Routing and queueing
 Wired routing, ad hoc routing and directed diffusion
 Queueing protocols: RED, drop-tail, etc.
 Physical media
 Wired (point-to-point, LANs), wireless (multiple propagation
models), satellite
 Tracing, visualization using Nam

8
 Ns
 Ns Programming
 Create the event scheduler
 Turn tracing
 Create network
 Setup routing
 Insert errors
 Create transport connection
 Create traffic Using Ns
(from http://www.isi.edu/nsnam/ns/ns-tutorial/)
 Transmit application-level
data

9
 Environments/Configurations
 set val(chan) Channel/WirelessChannel ;# channel type
 set val(prop) Propagation/TwoRayGround ;# radio-propagation model
 set val(ant) Antenna/OmniAntenna ;# Antenna type
 set val(ll) LL ;# Link layer type
 set val(ifq) Queue/DropTail/PriQueue ;# Interface queue type
 set val(ifqlen) 50 ;# max packet in ifq
 set val(netif) Phy/WirelessPhy ;# network interface type
 set val(mac) Mac/802_11 ;# MAC type
 set val(nn) 4 ;# number of mobilenodes
 set val(rp) AODV ;# routing protocol
 set val(x) 800
 set val(y) 800

10
 Simulation #1
 Scenario
 Two fixed nodes
 moving within 600m x 600m flat topology

 DSR ad hoc routing

 TCP and CBR traffic

 Receiver move in and out of range

 Results
 Time vs. packets arrived

11
 Simulation #2
 Scenario
 Two fixed pairs (4 nodes)
 moving within 800m x 800m flat topology

 AODV ad hoc routing

 TCP and CBR traffic

 2 nodes in each pair communicate each other

(hidden node)
 Results
 Time vs. packets arrived
12
 Simulation #3
 Scenario
 Six fixed nodes
 Change Routing algorithm

 4 Ad hoc routing: DSR/ DSDV/ AODV/ TORA

 The left-most node sends data to the right-most

node
 Results
 Time vs. packets arrived

13
 Further studies
 Check the effectiveness of RTS/CTS
handshake
 Consider a lot of nodes in a small space
 More experiments using other traffic model
(e.g. burst)
 Source-level (C++) modification for deeper
understanding

14
 Useful links
 Monarch project
 http://www.monarch.cs.rice.edu
 (more links will be added on the web)

15
That’s all

Thanks.

16