Transcript PPT - Suraj @ LUMS

Performance Evaluation of the
IEEE 802.16 MAC for QoS Support
Aemen Hassaan Lodhi
05060021
Multimedia
Communications Project
(Spring 2006-07)
Outline







IEEE 802.16 (WiMAX)
Project Objective
Simulation Environment
Performance Metrics
Experiments Carried out
Discussion of results
Conclusion
IEEE 802.16 (WiMAX)



Wireless MAN – provides network access to
subscriber stations (SS) with radio base
stations (BS)
Offers an alternative to cabled access
networks – fiber optic links, coaxial cables
using cable modems, DSL links
Supports nomadic and mobile clients on the go
(IEEE 802.16 e, 2004)
MAC layer







Manage the resources of the link air-link in efficiently
and provide Quality of Service (QoS)
differentiation for different connections/streams
Supporting Point to Multipoint and Mesh network
models
Performing Link Adaption & ARQ functions
Transmission Scheduling
Admission Control
Link Initialization
Fragmentation and Retransmission
Project Objective



Verify via simulation the ability of IEEE 802.16
MAC to handle different types of traffic having
different QoS requirements
The response of the protocol to different types
of traffic and environments
Simulations of IEEE 802.16 on NS-2
Simulation Environment





NS 2 Simulator
PMP mode
Time Division Multiple Access mode for
transmission from SSs to BS
Downlink and Uplink subframes duplexed
using Frequency Division Duplex
Full Duplex Subscriber Stations
Simulator environment
Simulator Environment
Problems with Chang Gung
University’s WiMAX Module



Couldn’t support more than 11~12 nodes
when actual simulations were carried out
Generated Segmentation faults
Even though the transmissions from source
nodes were cut-off the Base station
continued to send traffic to receiver nodes!!
Problems with Chang Gung
University’s WiMAX Module
WiMAX module by NIST, USA







Number of Nodes (the simulator crashes after 24 nodes)
Traffic (CBR, Poisson ON/OFF source, Pareto ON/OFF source
to model web traffic, Real Audio, any mix of the above.)
Radio Propagation models (2-Ray ground model for open/suburban areas and Shadowing model for urban areas)
Mobility of nodes
Packet Sizes
A number of parameters at the physical layer e.g. modulation
schemes
Does NOT offer choice of scheduling mechanism at the MAC
layer. Only Best Effort Scheduler is available
Traffic Models



Constant Bit Rate
Pareto ON/OFF source to model web traffic
Voice over IP traffic
–
–
–
–

Followed the specs of ITU G7.11
64Kbps during talk spurt (fixed)
Average length of talk spurt 352 ms
Average length of silence period 650 ms
Videoconference traffic could not be run
Performance Evaluation




Throughput for different types of traffic
%age utilization of bandwidth vs. Input load
Packet loss rate for different traffic streams
Comparison of the response of IEEE 802.16
with Best Effort Scheduler to different traffic
streams and different environments
Packet loss rate vs. Number of Nodes CBR
stream, Stationary nodes, 2Ray Ground
Packet loss rate vs. Number of Nodes Pareto
ON/OFF source, Stationary nodes, 2Ray Ground
Packet loss rate vs. Number of Nodes Pareto
ON/OFF stream, Stationary nodes, 2Ray
No. of mobile nodes
Packets Sent
Packets Received
1
24
22
3
72
66
5
120
110
7
168
154
9
216
192
11
264
242
13
312
286
15
360
323
17
408
368
20
480
429
24
576
516
Packet loss rate vs. Number of Nodes
CBR stream, Mobile nodes, 2Ray
Packet loss rate vs. Number of Nodes Pareto
ON/OFF stream, Mobile nodes, 2Ray
Packet loss rate vs. Number of Nodes CBR
stream, Stationary nodes, Shadowing
Packet loss rate vs. Number of Nodes Pareto
On/Off stream, Mobile nodes, Shadowing
Packet loss rate vs. Number of Nodes
VoIP Traffic, Stationary nodes, 2Ray
Packet loss rate vs. Number of Nodes
Traffic Mix (CBR + VoIP + Web)
CBR Stationary vs. Mobile Nodes
CBR, 2Ray Ground vs. Shadowing
Pareto On/Off source, 2Ray vs. Shadowing
CBR vs. Pareto throughput
Throughput (Mbps) vs. Number of
nodes
Percentage Utilization of Bandwidth vs.
Number of Nodes
Conclusions



Though utilization is high the standard with
its current specifications (70 Mbps) will
maximize its throughput at about 80 nodes.
Without a proper classification of streams at
the base station the streams with stringent
QoS will suffer great loss
With TDD polling mechanism the nodes
generating traffic with stringent QoS
requirements suffer a great deal
Packet loss rate vs. Number of Nodes
Traffic Mix (CBR + VoIP + Web)
References
1. Claudio Cicconetti, Alessandro Erta, Luciano Lenzini, and Enzo
Mingozzi, ‘Performance Evaluation of the IEEE 802.16 MAC for QoS
Support’, IEEE Transactions on Mobile Computing, VOL. 6, No. 1,
January 2007
2. Jenhui Chen , Chih-Chieh Wang, Frank Chee-Da Tsai§, Chiang-Wei
Chang, Syao-Syuan Liu, Jhenjhong Guo, Wei-Jen Lien, Jui-Hsiang
Sum, and Chih-Hsin Hung, ‘The Design and Implementation of WiMAX
Module for ns-2 Simulator’, ACM International Conference Proceeding
Series, Proceeding from the 2006 workshop on ns-2: the IP network
simulator
3. F.H.P. Fitzek and M. Reisslein, “MPEG4 and H.263 Video Traces for
Network Performance Evaluation,” IEEE Network Magazine, vol. 15,
no. 6, pp. 40-54 Nov. 2001.
4. NS-2 reference manual
5. ITU G7.11 specifications
References

“Radio Propagation Models”, Chapter 17, NS
Documentation

P.M. Fiorini, “Voice over IP (VoIP) for Enterprise
Networks: Performance Implications & Traffic
Models”.