J. Patrick Frantz Executive Director Center for Multimedia Communication Rice University IEEE Vehicular Technology Conference April 24, 2003 Performance of IEEE 802.11b Wireless LAN in an Emulated Mobile Channel
Jan 05, 2016
J. Patrick FrantzExecutive Director
Center for Multimedia CommunicationRice University
IEEE Vehicular Technology ConferenceApril 24, 2003
J. Patrick FrantzExecutive Director
Center for Multimedia CommunicationRice University
IEEE Vehicular Technology ConferenceApril 24, 2003
Performance of IEEE 802.11b Wireless LAN in an Emulated Mobile
Channel
Performance of IEEE 802.11b Wireless LAN in an Emulated Mobile
Channel
Performance of 802.11b is Well Understood for Indoor UseOutdoor and Mobile Environments Remain an Unexplored SpacePlans for Use in Such Environments Prompts Further Study (e.g. Mercedes & Other Telematics Applications)
Performance of 802.11b is Well Understood for Indoor UseOutdoor and Mobile Environments Remain an Unexplored SpacePlans for Use in Such Environments Prompts Further Study (e.g. Mercedes & Other Telematics Applications)
MotivationMotivation
Reviewed Literature for Measurements of Outdoor 2.4GHz ISM ChannelDetermined a Reasonable Set of Channel Parameters for Repeatable ExperimentsPrecise Variation of Parameters to Measure Relationship to System PerformanceUsed Channel Emulation for Reliable & Repeatable Results
Reviewed Literature for Measurements of Outdoor 2.4GHz ISM ChannelDetermined a Reasonable Set of Channel Parameters for Repeatable ExperimentsPrecise Variation of Parameters to Measure Relationship to System PerformanceUsed Channel Emulation for Reliable & Repeatable Results
MethodologyMethodology
Experimental SetupExperimental Setup
Setup 3 Peer-to-Peer Links With H/W From 2 Different Manufacturers
Wanted to Test Differences in Baseband Implementations from Two Leading Chipset Manufactureres
AgereIntersil
Setup 3 Peer-to-Peer Links With H/W From 2 Different Manufacturers
Wanted to Test Differences in Baseband Implementations from Two Leading Chipset Manufactureres
AgereIntersil
2 Orinoco WLAN PC Cards (Agere Chipset)2 D-Link Air DWL-520 WLAN PCI Cards (Intersil Chipset)TAS 4500 FLEX5 RF Channel Emulator from Spirent CommunicationsIPERF v1.1.1 Network Testing Software
2 Orinoco WLAN PC Cards (Agere Chipset)2 D-Link Air DWL-520 WLAN PCI Cards (Intersil Chipset)TAS 4500 FLEX5 RF Channel Emulator from Spirent CommunicationsIPERF v1.1.1 Network Testing Software
Equipment UsedEquipment Used
Experimental SetupExperimental Setup
MethodologyMethodology
Velocity - Combined Effect of Ricean and Rayleigh Fast Fading with Doppler ShiftDelay Spread - Created an Ad Hoc Multipath Channel to Study Effects of Delay Spread (RMS)
Similar to Power Delay Profiles from LiteratureUsed 6-Path Model
Velocity - Combined Effect of Ricean and Rayleigh Fast Fading with Doppler ShiftDelay Spread - Created an Ad Hoc Multipath Channel to Study Effects of Delay Spread (RMS)
Similar to Power Delay Profiles from LiteratureUsed 6-Path Model
Variable Modulations - Constrained the Hardware to Specific Modulations (1, 2, 5.5 & 11Mbps)SNR - Monitored and Recorded SNR at both ends of Test Setup to Account for Path Losses in the System.Careful System Isolation and Shielding
Variable Modulations - Constrained the Hardware to Specific Modulations (1, 2, 5.5 & 11Mbps)SNR - Monitored and Recorded SNR at both ends of Test Setup to Account for Path Losses in the System.Careful System Isolation and Shielding
MethodologyMethodology
Both Implementations of 802.11b Showed Significant Performance Degredations in Mobile Channels, as ExpectedStrong Link Between SNR and Flat Fading & Delay Spread Susceptibility
Both Implementations of 802.11b Showed Significant Performance Degredations in Mobile Channels, as ExpectedStrong Link Between SNR and Flat Fading & Delay Spread Susceptibility
Experimental ResultsExperimental Results
Throughput vs. Velocity in Rayleigh & Ricean Fading
Throughput vs. Velocity in Rayleigh & Ricean Fading
Packet Loss vs. Velocity in Rayleigh & Ricean Fading
Packet Loss vs. Velocity in Rayleigh & Ricean Fading
Throughput and Packet Loss vs. Delay Spread
Throughput and Packet Loss vs. Delay Spread
Throughput & Packet Loss vs. Velocity (D-Link, Rayleigh, 15dB)
Throughput & Packet Loss vs. Velocity (D-Link, Rayleigh, 15dB)
Throughput & Packet Loss vs. Velocity (D-Link, Rayleigh, 15dB)
Throughput & Packet Loss vs. Velocity (D-Link, Rayleigh, 15dB)
Throughput & Packet Loss vs. Delay Spread
Throughput & Packet Loss vs. Delay Spread
ConclusionsConclusions
Typical Implementations of 802.11b are Far Less Effective in Mobile ChannelHowever, No Reason that Current Implementations Can’t be ModifiedRAKE Receivers for Indoor Channels are Insufficient for Outdoor Channels with High RMS Delay SpreadSlight Incompatibilities Between Chipsets Adversely Affect Performance
Typical Implementations of 802.11b are Far Less Effective in Mobile ChannelHowever, No Reason that Current Implementations Can’t be ModifiedRAKE Receivers for Indoor Channels are Insufficient for Outdoor Channels with High RMS Delay SpreadSlight Incompatibilities Between Chipsets Adversely Affect Performance