Rate Mutilation Improvement for Cross section based P2P Video Gushing.


79 views
Uploaded on:
Description
Moderator: Dr. Sachin Agarwal. Deutsche Telekom Laboratories. Rate Distortion Optimization ... Cross section based P2P can completely use the system assets of its associates ...
Transcripts
Slide 1

Rate Distortion Optimization for Mesh-based P2P Video Streaming Tareq Hossain, Yi Cui, Yuan Xue V anderbilt A dvanced Net work and S ystems Group Vanderbilt University, USA Presenter: Dr. Sachin Agarwal Deutsche Telekom Laboratories

Slide 2

Outline Motivation Video Broadcast Can P2P Help? Rate Distortion for P2P Mesh Rate Optimization Simulation Results Conclusion

Slide 3

Motivation Video Broadcast Increasing prevalence because of wide utilization of web Can P2P Help? Savvy asset use CPU cycles Storage space Uplink data transmission Instant deployability Almost pervasive system scope without CDN administrations and IP multicast

Slide 4

Rate Distortion for P2P Mesh based P2P can completely use the system assets of its associates contrasted with a tree based system We utilize circulated calculation – every companion modifies its own gushing rate to achieve the worldwide ideal by fulfilling: Capacity imperative Relay limitation Double Pricing Solution Simultaneous consolidation of limit requirement and transfer imperative altogether diminishes the total rate mutilation Single Pricing Solution Relay imperative is connected after rate contortion calculation focalizes We show rate-bending advancement for P2P network arrange Double estimating arrangement performs superior to anything single valuing arrangement

Slide 5

Outline Motivation Rate Optimization Performance Evaluation Problem Formulation Distributed Algorithm Simulation Results Conclusion

Slide 6

Performance Evaluation Video quality is measured as the Mean-Square-Error (MSE) found the middle value of over all casings PSNR is utilized to evaluate video quality, characterized by D speaks to the general Mean-Square-Error (MSE) arrived at the midpoint of over all edges of an encoded video succession The twisting D as an element of spilling rate x f is given by The variables ( θ , x 0 and D 0 ) rely on upon encoded video grouping and additionally on the rate of intra coded macroblocks.

Slide 7

Problem Formulation Rate improvement is a curved capacity of the distributed rate Here f speaks to a stream between two associates, x is the rate vector and c is the limit vector An is a L x F (join, stream) grid of connections and streams such that A lf = 1 if stream f experiences join l and 0 generally B is a F x F meager lattice, where (( h k – 1 ) H + h i )th line is dynamic just if there is a stream from companion h k to companion h i . Formally,

Slide 8

Distributed Algorithm Each getting peer ( ) figures the rates of its approaching streams in a cross section Network value: Net transfer value: Source Rate upgrade for every companion: Rate is overhauled in light of the base of system and net hand-off cost accessible among the all approaching streams Rate redesign for approaching streams: Rate upgrade for approaching streams with least system and net hand-off value: join value hand-off cost

Slide 9

Outline Motivation Rate Optimization Simulation Configuration Input Data Multicast Tree Construction Results Conclusion

Slide 10

Configuration To decide the genuine designated rate, we pick the most noteworthy quantized rate that is instantly not exactly the rate accomplished by our answer The ITU-T test arrangements utilized are: foreman, akiyo, corridor, mother-little girl T he server has a settled rate of 2Mbps The greatest number of associates ~160 The uplink transmission capacity of every companion is haphazardly appointed somewhere around 0.6Mbps and 2Mbps

Slide 11

Input information The PSNR-Rate video information (an) and Number of associates Time information (b): Rate (Kbps)

Slide 12

Multicast Mesh Construction Peers join the gushing system one-by-one Joining peer utilizes the extra limit of existing companions to decide a reasonable guardian. The extra coefficient is characterized as Here x f(h) is the approaching stream rate of the associate h Implementation At the end of every rate redesign cycle, peers send their extra coefficient worth to guardians The ID of the best reasonable guardian engenders to the server

Slide 13

Outline Motivation Rate Optimization Simulation Results Conclusion

Slide 14

Results The normal PSNR increase over every one of the recordings for the twofold evaluating arrangement is 1.86 dB (PSNR is 0 when all companions leave ~720s)

Slide 15

Results The normal addition for the twofold estimating arrangement spoke to as far as rate

Slide 16

Conclusion We display an ideal rate designation answer for P2P network system We utilize non-direct improvement structure Minimize total bending Maximize the general PSNR among all companions in a P2P work Simultaneously apply peer handing-off imperative alongside limit requirement Double valuing arrangement reliably performs superior to anything single valuing arrangement

Slide 17

Thank You VANETS (Vanderbilt Advanced Network and Systems) Group http://vanets.vuse.vanderbilt.edu QUESTIONS?

Recommended
View more...