2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698) 2003
DOI: 10.1109/icme.2003.1221667
|View full text |Cite
|
Sign up to set email alerts
|

Comparison of rate allocation strategies for H.264 video transmission over wireless lossy correlated networks

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1

Citation Types

0
5
0

Year Published

2004
2004
2007
2007

Publication Types

Select...
5
3

Relationship

1
7

Authors

Journals

citations
Cited by 11 publications
(5 citation statements)
references
References 5 publications
0
5
0
Order By: Relevance
“…4, we illustrate a plot of PSNR versus the packet loss rate with for the Foreman sequence at , which is in the range of low bit-rate services. The average burst length is chosen to be as described in [14], [25] in order to provide a representative burst-length distribution as expected in a typical UMTS network. The proposed adaptive scheme can be seen to achieve a much higher performance in terms of end-to-end PSNR compared to the representative nonadaptive schemes which apply the RS (15,9) code to every frame, together with fixed intra-updating rates, regardless of its motion level and other source/channel conditions.…”
Section: B Selected Simulation Results and Discussionmentioning
confidence: 99%
“…4, we illustrate a plot of PSNR versus the packet loss rate with for the Foreman sequence at , which is in the range of low bit-rate services. The average burst length is chosen to be as described in [14], [25] in order to provide a representative burst-length distribution as expected in a typical UMTS network. The proposed adaptive scheme can be seen to achieve a much higher performance in terms of end-to-end PSNR compared to the representative nonadaptive schemes which apply the RS (15,9) code to every frame, together with fixed intra-updating rates, regardless of its motion level and other source/channel conditions.…”
Section: B Selected Simulation Results and Discussionmentioning
confidence: 99%
“…1 reports the average burst length of packet errors, as a function of packet payload size (i.e., the size of each slice). As can be seen, at typical packet sizes (e.g., from 500 to 1500 bytes) the average burst length is always larger than one, and typically lies between 1.5 and 2.5 [25]. This indicates that not only the event of losing one complete frame is most likely, but one should also take into account the possibility that two (or more) consecutive frames are lost.…”
Section: B Motivationmentioning
confidence: 94%
“…The response packet we use is 5bytes long including 32 bits packet index and 8 bits response type information. Refers to the network interference to data packet and response packet, the burst error affects the smaller packet, as response packet [14] , much more, while the packet loss rate is higher to longer packet, as data packet, in the same bit error rate environment [8] .…”
Section: ) Timingmentioning
confidence: 98%
“…simulate a more reliable, lower error-rate bearer that is required in conversational applications [8] . We choose pattern 1 in our parameter optimization simulation and set the corresponding packet loss rate and burst errors in [8] and [14] for data transmission packet and the response packet depending on the different packet size.…”
Section: ) Timingmentioning
confidence: 99%
See 1 more Smart Citation