a cooperative multicast scheduling scheme for multicast services in ieee 802.16 networks fen hou,...
TRANSCRIPT
A Cooperative Multicast Scheduling Scheme for A Cooperative Multicast Scheduling Scheme for Multicast Services in IEEE 802.16 NetworksMulticast Services in IEEE 802.16 Networks
Fen Hou, Pin-Han Ho, and Lin X CaiFen Hou, Pin-Han Ho, and Lin X Cai
Dept. of Electrical and Computer Engineering, Waterloo UniversityDept. of Electrical and Computer Engineering, Waterloo University
IEEE Transactions on Wireless Communications, Vol. 8, No 3, March, 2009IEEE Transactions on Wireless Communications, Vol. 8, No 3, March, 2009
22
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
AgendaAgenda
IntroductionIntroduction Cooperative Multicast Scheduling SchemeCooperative Multicast Scheduling Scheme Performance AnalysisPerformance Analysis Simulation ResultsSimulation Results ConclusionsConclusions
33
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
IntroductionIntroduction
Multicast communicationMulticast communication An efficient mechanism for one-to-many transmissions ovAn efficient mechanism for one-to-many transmissions ov
er wireless channeler wireless channel Users’ requests are the sameUsers’ requests are the same
Multicast group (MGroup)Multicast group (MGroup) ApplicationsApplications
Emergency alert Emergency alert Internet Protocol Television (IPTV)Internet Protocol Television (IPTV) Mobile TV Mobile TV
44
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
IntroductionIntroduction-Motivation-Motivation
MGroupsMGroups
Lag !!
Lag !!
Lag !!
Lag !!
55
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
IntroductionIntroduction-Motivation-Motivation
Scheduling mechanismScheduling mechanism Fairness considerationFairness consideration
Round-robin for each MgroupRound-robin for each Mgroup Ex. CDMA 2000 1xEx-Do networks : 204.8 kbpsEx. CDMA 2000 1xEx-Do networks : 204.8 kbps
Robust considerationRobust consideration Minimal supported rate of all Mgroup membersMinimal supported rate of all Mgroup members
Group member with the worst channel condition becomes bottlenGroup member with the worst channel condition becomes bottleneckeck
Inefficient to most users in good channel conditionsInefficient to most users in good channel conditions
66
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
IntroductionIntroduction-Goal-Goal
Cooperative multicast communicationsCooperative multicast communications Achieve high throughput for all group membersAchieve high throughput for all group members Maintain good fairness performanceMaintain good fairness performance
77
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Cooperative Multicast Scheduling SchemeCooperative Multicast Scheduling Scheme
Multicast group selectionMulticast group selection Decide which groups should be served at each frameDecide which groups should be served at each frame
Cooperative multicast transmissionCooperative multicast transmission Decide how to multicast data to the selected group Decide how to multicast data to the selected group
membersmembers
88
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Multicast group selectionMulticast group selection
Random MGroup SelectionRandom MGroup Selection BS randomly selects an MGroup for service with a pre-defBS randomly selects an MGroup for service with a pre-def
ined probabilityined probability The same probability for each MGroupThe same probability for each MGroup Service differentiation Service differentiation
99
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Multicast group selectionMulticast group selection
Channel-aware MGroup SelectionChannel-aware MGroup Selection Normalized relative channel condition Normalized relative channel condition XXi i for MGroupsfor MGroups
Normalized relative channel condition of MGroup i
Total number of group members in MGroup i
Instantaneous SNR of the j-th group members in MGroup i
Avg. SNR of the j-th group members in MGroup iThe set of all group mem
bers in MGroup i
1010
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Cooperative Multicast TransmissionCooperative Multicast Transmission
DilemmaDilemma
64-QAM
16-QAM
QPSK
64-QAM
16-QAM
QPSK
64-QAM QPSKToo fast !!
Too fast !!
Oh! Cool! Too
Slow !!Too
Slow !!
Oh! Cool!
1111
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Cooperative Multicast TransmissionCooperative Multicast Transmission
Two-phase transmission schemeTwo-phase transmission scheme
Downlink Burst # i
Phase 1 ( ) 2iR Phase 1 ( ) 2
iR
64-QAM 16-QAM
1212
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Cooperative Multicast TransmissionCooperative Multicast Transmission
AdvantagesAdvantages Higher throughput than conservative rateHigher throughput than conservative rate
Group members with good channel conditionsGroup members with good channel conditions Group members with bad channel conditionsGroup members with bad channel conditions
ExtensionsExtensions m-m-phase transmissionsphase transmissions Larger Larger mm, lager computation and control overhead, lager computation and control overhead
1313
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Performance AnalysisPerformance Analysis
Service probability of each MGroupService probability of each MGroup Throughput of each user Throughput of each user Throughput of each MGroupThroughput of each MGroup Throughput of the whole networksThroughput of the whole networks
1414
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Service Probability of Each MGroupService Probability of Each MGroup
DefinitionDefinition The probability than an MGroup is selected for service in The probability than an MGroup is selected for service in
a MAC frame when the system is stablea MAC frame when the system is stable
1515
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Throughput of each MGroupThroughput of each MGroup
Given and , the probability that a group meGiven and , the probability that a group member in MGroup mber in MGroup i i , SS, SSii,,j j can receive the data in Phacan receive the data in Phase 1se 1
)1(log)1(log0
1,
221
N
ECSNRCR ji
i
)1(220
1,1
N
E jiCRi 01. )12(
1
NE iRji
jBiiR
iiiENSNRSNRR
ji eeNE ,0
11 /))12((/
01, ))12(Pr( jBi
iRiii
ENSNRSNRRji eeNE ,0
11 /))12((/
01, ))12(Pr(
(Shannon Capacity)
The avg. received power for SSi,j from BS
1616
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Throughput of each MGroupThroughput of each MGroup
Given and , the probability that a group meGiven and , the probability that a group member in MGroup mber in MGroup i i , the set of all possible for SSi,, the set of all possible for SSi,j j
121
||1
1
1,
i
i
N
k
Niji k
NC
giG
Example: Suppose MGroup i contains SSi,1, SSi,2, SSi,3, SSi,4. The possible group members that can be transmitter in phase 2 are SSi,2, SSi,3, SSi,4.The number of possible are
giG
3
1, 7
3||
kji k
C
SSi,2 SSi,3 SSi,4
SSi,2 SSi,3 SSi,2 SSi,4SSi,3 SSi,4
SSi,2 SSi,3 SSi,4
1717
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Throughput of each MGroupThroughput of each MGroup
Given and , the probability that a group meGiven and , the probability that a group member in MGroup mber in MGroup i i , SS, SSii,,j j cannot receive the data in Pcannot receive the data in Phase 1, but can receive the data in Phase 2hase 1, but can receive the data in Phase 2
1818
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Throughput for SSThroughput for SSi,ji,j
Throughput of each MGroupThroughput of each MGroup
Throughput of each MGroupThroughput of each MGroupPhase 1 Successful
Phase 1 Failure, Phase 2 Successful
1919
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Simulation ResultsSimulation Results
2020
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Simulation ResultsSimulation Results
2121
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Simulation ResultsSimulation Results
2222
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Simulation ResultsSimulation Results
2323
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
Simulation ResultsSimulation Results
2424
Speaker : Chi-Tao Chiang (蔣季陶 )
Meeting 報告簡報
ConclusionConclusion
Cooperative multicast scheduling scheme for multimCooperative multicast scheduling scheme for multimedia services in IEEE 802.16 networksedia services in IEEE 802.16 networks
Simulation results shows that Simulation results shows that Analysis and the proposed scheme are similar in terms ofAnalysis and the proposed scheme are similar in terms of
Throughput Throughput Service probabilityService probability
Higher throughput Higher throughput