M EDUSA : a Reliable Chunkless Peer-to-Peer Architecture for - - PowerPoint PPT Presentation

m edusa a reliable chunkless peer to peer architecture
SMART_READER_LITE
LIVE PREVIEW

M EDUSA : a Reliable Chunkless Peer-to-Peer Architecture for - - PowerPoint PPT Presentation

M EDUSA : a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming R. Bernardini and R. Rinaldo and A. Vitali DIEGM University of Udine, Udine, Italy ST microelectronics, Agate Brianza, Italy M EDUSA : a


slide-1
SLIDE 1

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

  • R. Bernardini∗ and R. Rinaldo∗ and A. Vitali†

∗ DIEGM – University of Udine, Udine, Italy † ST microelectronics, Agate Brianza, Italy

slide-2
SLIDE 2

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Outline

  • Introduction

– Why P2P live streaming? – Problems

  • Our proposal
  • Conclusions

1

slide-3
SLIDE 3

Introduction

slide-4
SLIDE 4

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Our objective

A peer-to-peer architecture for video live streaming

2

slide-5
SLIDE 5

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Applicative context

group #4 User

Sub−source #4

Root Source

group #3 User

Sub−source #3 Sub−source #2

group #2 User

1 3

Dispatcher

Sub−source #1 2

group #1 User

New User Live Event

Hi, some music, please Contact user H i n e w u s e r , w e l c

  • m

e t

  • g

r

  • u

p # 1

3

slide-6
SLIDE 6

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Why peer-to-peer?

Bandwidth # Users

With the centralized approach the bandwidth required to each server grows linearly with the number of users With the P2P approach, each new user is also a new server ⇒ enhanced scalability

4

slide-7
SLIDE 7

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Obstacles to P2P live streaming

5

slide-8
SLIDE 8

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Obstacles to P2P live streaming (2)

Evanescence

Peer Peer Peer Peer Peer Peer Peer Peer Peer Peer Peer Bye...

? ?

Server

  • Peers can leave at any time
  • Nodes can remain isolated

⇒ quality drop

  • Solution:

– Network structure – Cross-packet FEC

6

slide-9
SLIDE 9

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Obstacles to P2P live streaming (3)

Asymmetric bandwidth

Download

Upload

Home users have enough download bandwidth to receive streaming data, . . . but. . . they have not enough upload bandwidth to act as streaming servers

Clipart courtesy FCIT http://etc.usf.edu/clipart

7

slide-10
SLIDE 10

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Obstacles to P2P live streaming (4)

Asymmetric bandwidth: the chunky solution

Chunk 4 Peer 4 Peer 1 Chunk 1 Chunk 2 Peer 2 Peer 3 Chunk 3 Me

File

Typical P2P approach: the file is partitioned into chunks, different peers send me different chunks.

8

slide-11
SLIDE 11

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Obstacles to P2P live streaming (5)

Drawbacks of chunky

  • The chunky solution works fine for “normal” files, but. . .

. . . it is not much useful with live streaming because – Future chunks do not exist yet – Past chunks are useless

  • Possibile solution: buffering

⇒ long start-up times

9

slide-12
SLIDE 12

Proposed solution

Description

slide-13
SLIDE 13

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

The problem

  • We want to break the original stream into smaller pieces (called crumbs)

such that – The bandwidth required by a crumb is N times smaller than the band- width required by the original stream – It is possible to recover the original stream from any combination of N different crumbs

New crumb Recovered packet

Me

Crumb

10

slide-14
SLIDE 14

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

The solution (. . . and secret sharing let it be. . . )

  • Work at packet level
  • A packet can be though as vector

c = [c1,c2,...,cK]t, ci ∈ GF(2n)

  • Vector c is organized as an N-row matrix and left

multiplied by vector r(b) := [1,b,...,bN−1] u := r(b)C = r(b)     c1 cN+1 ··· c2 cN+2 ··· . . . . . . cN c2N ···    

  • b and u are sent to other peers

r: reduction vector N: reduction factor u: reduced vector

11

slide-15
SLIDE 15

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Reconstruction HOWTO

  • Original packet can be reconstructed as soon as N crumbs are received.
  • Reconstruction is carried out by solving linear system

           u1 = r1C u2 = r2C . . . = . . . . . . uN = rNC − → C =     r1 r2 . . . rN    

−1

  • R−1

    u1 u2 . . . uN    

  • U

12

slide-16
SLIDE 16

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Reconstruction HOWTO (2)

  • The matrix of reduction vectors

R =     r1 r2 . . . rN     =      1 b1 b2

1

··· bN−1

1

1 b2 b2

2

··· bN−1

2

. . . . . . . . . . . . 1 bN b2

N

··· bN−1

N

     is a Wandermonde matrix ⇒ Invertibility is granted as soon as bi = b j for every i = j

13

slide-17
SLIDE 17

Proposed solution

Remarks

slide-18
SLIDE 18

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Characteristics of proposed solution

Independent on the P2P structure Independent on the multimedia format. It can be used with – Scalable streams – Multiple description – Encrypted format

14

slide-19
SLIDE 19

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Choosing the reduction vector

  • The reduction vector is chosen as

r(b) = [1, b, b2, ..., bN−1] where b is a random non-null element of GF(2n).

  • Random choice ⇒ simpler structure

⇒ R maybe non full-rank

15

slide-20
SLIDE 20

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Choosing the reduction vector (2)

Probability of non-full-rank R

10 20 30 40 50 10

−10

10

−8

10

−6

10

−4

10

−2

10 N Prob. F28 F216 F232

Birthday problem

16

slide-21
SLIDE 21

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Failure probability

  • To combact evenescence we can contact M > N peers.
  • The number P of currently available peers can be modeled as a queue with

M servers and capacity M

  • We experience “failure” (we are unable to reconstruct the signal) if P < N

2 1 M N N−1 N+1 n n+1 State Number of connected peers

GOOD

1/T find

conn

(n+1)/T

Tconn/Tfind

New peer found Peer leaves

BAD

ρ=

17

slide-22
SLIDE 22

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Failure probability (2)

0.2 0.4 0.6 0.8 1 1.2 1.4 10

−10

10

−8

10

−6

10

−4

10

−2

10 M/ρ PF α=M/N=1.10 ρ=50 ρ=100 ρ=200 ρ=400 ρ=800 0.6 0.7 0.8 0.9 1 1.1 1.2 10

−10

10

−8

10

−6

10

−4

10

−2

10 M/ρ PF ρ=400.0 α=1.1 α=1.2 α=1.3 α=1.5 α=2.0

18

slide-23
SLIDE 23

Extensions

slide-24
SLIDE 24

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Unequal reduction

Glossary

Base Layer Stream Detail Layer

Packet Analog Analog Binary Packet Analog Analog Binary Packet Analog Analog Binary Packet Analog Analog Binary

...

Packet Analog Analog Binary Packet Analog Analog Binary Packet Analog Analog Binary Packet Analog Analog Binary

...

Multimedia content Stream

  • The multimedia content (e.g., a soccer

match) is a colletion one or more streams (e.g., audio/video, different SVC layers, dif- ferent descriptions, . . . )

  • Each stream is a sequence of multimedia

packets (e.g., data relative to a slice).

  • Each multimedia packet can further di-

vided into different components (e.g., motion vectors, DCT coefficients, head-

  • ers. . . ).

Reduction is done at the component level

19

slide-25
SLIDE 25

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Unequal reduction (2)

Crumb creation

Reduced 1 Reduced 2 Reduced 3 Reduced 1 Reduced 2 Reduced 3

Reduction Reduction Reduction

Packet

Crumb

Component 3 Component 2 Component 1 Component 2 Component 3 Component 1

20

slide-26
SLIDE 26

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Digital Right Management

  • In some context, only authorized users should be able to access the multi-

media content

  • A possible solution is encrypting the data stream
  • For increased security, decryption should be done inside the player, not in-

side the P2P software Encryption must be transparent to the P2P software

21

slide-27
SLIDE 27

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Seeder with cryptography

Video archive

Parser Content Key Binary/Analog Components Reducer Reducer Reducer Encryption Encryption P2P Module To the peers To/from user Public user key

  • For

every content the seeder generates a content key

  • The content key is used

to encrypt the analog/binary components

  • When a new node joins

the network it gives to the seeder the public key of its player

  • The

seeder encrypts the content key with the pub- lic key and sends the result back to the user 22

slide-28
SLIDE 28

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Node with cryptography

Player Decryption Deparser Decryption P2P Module

MD module

Virtual player

Private user key Public user key

To/from seeder From other peers To other peers

Content key

  • To

join the network the nodes sends the public key

  • f its virtual player to the

seeder (via the P2P module)

  • The P2P module receives

the encrypted content key and gives it to the virtual player

  • The virtual player obtains

the content key by using its private key

  • The content key is used to

decrypt the components. 23

slide-29
SLIDE 29

Conclusions

slide-30
SLIDE 30

MEDUSA: a Reliable Chunkless Peer-to-Peer Architecture for Multimedia Streaming

Conclusions

  • We presented a chunkless approach to live video distribution over P2P
  • The proposed approach

– solves both the asymmetric bandwidth and the evanescence problems. – is independent on the multimedia format – is transparent with respect to DRM techniques

24