1. Scalable On-Demand Media Streaming With Packet Loss Recovery Anirban Mahanti, Derek L. Eager, Mary K. Vernon, and David J. Sundaram-Stukel IEEE/ACM Trans. on Networking, April 2003

2. Outline Introduction Periodic broadcast Bandwidth skimming Lower bound (server bandwidth) Loss recovery strategies Erasure code Retransmission Reliable periodic broadcast

3. Periodic broadcast (startup delay) Divide the media file into many segments. Each segment is repeatedly broadcast. Ex: Harmonic, Pyramid, Skyscraper, …

4. Bandwidth skimming (no delay) Initiate a multicast stream for each request. Merge streams to reduce network bandwidth.

5. Notations

6. Maximum achievable scalability (1/2) The lower bound is derived on the required server bandwidth for any protocol that provides immediate on-demand streaming of multimedia content. Consider a small portion of the object at time offset x: For an client request that arrives at time t, this portion of the object must be delivered no later than time t+x. If the portion is multicast at time t+x, then (at best) those clients that request the object from time t and t+x can receive the same multicast.

7. Maximum achievable scalability (2/2) Assuming Poisson arrivals, the average time from t+x until the next request of the object is 1/. The minimum frequency of multicasts of the portion at time offset x is 1/(x+1/ ). The lower bound is tt+xNext request 1/

8. Required server bandwidth (d=0)

9. Required server bandwidth (d>0) The bound can be extended by adding a startup delay d.

10. Loss recovery strategies (1/2) Multicast transmission Use erasure codes. If the average client packet loss probability is p, the lower bound for immediate service using erasure codes is The lower bound for streaming with bounded delay is

11. Loss recovery strategies (2/2) Unicast transmission Use retransmission. If the object playback duration is T, the average amount of data that is retransmitted per client is pT/(1-p). Given a client request rate, the server bandwidth required for retransmitted data is pT/(1-p). Because N= T, the minimal server bandwidth is

12. B min (15% packet loss)

13. Reliable periodic broadcast Optimized PB protocols Have the minimum possible startup delay. (under their model) Do not support packet loss recovery. Basic RPB Using erasure codes. RPB protocols for bursty packet loss Allow each segment to have a different associated cumulative loss protection.

14. Parameters

15. Optimized PB protocols (1/4) Assume a maximum aggregate transmission rate to any given client is b. Each segment must be entirely received before the beginning of the segment is played. Assume r=1, For each segment k, 1<k  s, it has maximum l k equal to (the time to receive segment 1 + the time to play segment 1 ~ k-1).

16. Optimized PB protocols (2/4) s = 3 1 2 3 4 5 k > s

17. Optimized PB protocols (3/4) For segment k>s, the client will begin receiving segment k at the time that segment k-s is just received and starts playing. The client must finish receiving segment k by the time that segment k-1 finishes playing. If l 1 =1, the maximum sizes for segment k>1:

18. Optimized PB protocols (4/4) If we use K server streams to multicast the object, The total server bandwidth B=r  K. The client startup delay (time to receive segment 1) =T/(r  l k ) K = 6, r = 1, s = 2

19. Required server bandwidth b = s  r

20. Multicast frequency The lower bound: the portion of the object at position x must be broadcast with frequency at least 1/(x+d).

21. Required client buffer space

22. Basic RPB Each segment is encoded using an erasure code. A client can listen to each channel until it has correctly received the number of packets required to reconstruct the respective segment. (1-p) Let a=1/(1-p), the maximum relative segment sizes are

23. Example Optimized RPB protocol K = 6, r = 1, s = 2, 10% packet loss No boundary!! (because of erasure code)

24. Performance (3% loss)

25. Performance (10% loss)

26. Performance (25% loss)

27. Impact of segment streaming rate

28. RPB protocols for bursty packet loss (1/2) If a given client observes cumulative packet rate less than p at the end of receiving a given segment, that segment can be reconstructed before its playout point. In this case, the client can begin listening to later segments earlier. This “ work-ahead ” allows the client to tolerate a higher loss rate than p for a later segment.

29. RPB protocols for bursty packet loss (2/2) Packet loss is bursty, so earlier segments require a higher level of loss protection than later segments. Assume each segment has a different a k, the segment sizes for the specified values of a k are:

30. Performance