Abstract
Reliable real time wireless video streaming is a challenging network application due its high bandwidth and low delay/jitter requirements. IEEE 802.11e helps address such QoS challenges, e.g. by prioritising video over background traffic. The reliability problem occurs since it is very challenging (near impossible) for a WLAN protocol to guarantee QoS, especially with uncoordinated contention being the main access mechanism, along with increasing numbers of stations and interfering traffic, with the hostile wireless radio
environment making any guarantee foolhardy. With the use of RTP, the video application has some control over end-to-end delay and jitter. Still, in the presence of excessive interference, RTP packet dropping might result in video quality degradation or video outage time. We study the reliability of wireless RTP video streaming over the EDCA MAC protocol, with the ability to forecast (and mitigate the effect of) video failures. Specifically, we introduce a model to evaluate both steady-state and transient video failures due to RTP buffer overflow, the overflow in turn due to WLAN co-channel interference. Quantitative reliability measurements are taken by configuring RTP and EDCA. We evaluate the average packet loss and average outage duration and we validate the theoretical results in an experimental setup. The proposed forecast mechanism could be coupled with and help optimise a) WLAN channel assignment in an ESS network, b) multipath (multi-channel/radio/hop) video routing, and c) adaptive rate control.
environment making any guarantee foolhardy. With the use of RTP, the video application has some control over end-to-end delay and jitter. Still, in the presence of excessive interference, RTP packet dropping might result in video quality degradation or video outage time. We study the reliability of wireless RTP video streaming over the EDCA MAC protocol, with the ability to forecast (and mitigate the effect of) video failures. Specifically, we introduce a model to evaluate both steady-state and transient video failures due to RTP buffer overflow, the overflow in turn due to WLAN co-channel interference. Quantitative reliability measurements are taken by configuring RTP and EDCA. We evaluate the average packet loss and average outage duration and we validate the theoretical results in an experimental setup. The proposed forecast mechanism could be coupled with and help optimise a) WLAN channel assignment in an ESS network, b) multipath (multi-channel/radio/hop) video routing, and c) adaptive rate control.
| Original language | English |
|---|---|
| Title of host publication | 7th ACM Workshop on Performance Monitoring and Measurement of Heterogeneous Wireless and Wired Networks |
| Subtitle of host publication | ACM MSWIM 2012 |
| Place of Publication | Cyprus |
| Pages | 149-185 |
| DOIs | |
| Publication status | Published - 21 Oct 2012 |