1. Active Networking and End-to-End Argument Samrat Bhattacharjee Kenneth L. Calvert Ellen W. Zegura

2. Objective End-to-End Argument Active Networking – Extension of E2E argument End System approach Vs Combined System approach Performance Model to quantify the above approaches

3. E2E and Active Networking What is E2E? Is Active Networking a natural consequence of E2E? E2E and placement of functionality  All applications might not use the service  Trade-off between performance and cost  Combine network and application information to optimize performance

4. Model for Service Location Analyze performance under two design options –  Design X : Service implementation exclusive in the end-systems  Design C : Service achieved through combination of implementation at the end-system and in the network Network is treated monolithically Network support is boolean

5. Model for Service Location Parameters of performance model  Exclusively End-system (Design X) Tx – Expected Performance  Combined End-system and Network (Design C) Tc – Expected performance Pn – Probability that the network support accomplishes the service Te – Expected performance, end-system version Tn – Expected performance, network version Tc = (1-Pn) Te + Pn Tn

6. Reliable Data Transfer Performance Metric : Expected Transfer Time Design X :  tx – time to request, receive and check the integrity  p – probability of error in each transmission  Tx – expected transfer time Tx =   i=1 P(i transmissions) * i * tx = tx / (1-p)

7. Reliable Data Transfer... Design C  tc – time to request, receive and check the integrity  p – probability of error in each transmission  q – probability that the network can correct the error Pn Tn = (1 – p + pq) tc Te = tc +   i=1 P(i transmissions) * i * tc = tc (1+1 / (1-p+pq) ) Tc = (1 – p + pq) tc + p(1-q) * tc (1+1 / (1-p+pq) )

8. Reliable Multicast Performance metric : Latency (no of hops) Design X  Buffering and Retransmission done only at the Receivers  Request message is directed to a “nearby” Receiver through the Loss node Design C  Buffering and Retransmission done by the network nodes

9. Reliable Multicast...

10. T x = t R + t L + t Y + 2 t R’ + t E + t R T N = 2(t R + t L + 1) T E = 2(t R + t L + t L ’ + t S ) T c = 2p n (t R + t L + 1) + 2 (1-p n ) (t R + t L + t L ’ + t S ) assume tR = tR’ = tS and tL = tL’ T x = 4t R + t L + t Y + t E T c = 4t R + 4t L - 2p n (t R + t L - 1) If p n > (3t L - t E - t Y ) / (2(t R + t L - 1)) then T c < T x ( combined system approach is better)

11. Congestion Control Application knows how to adapt Network knows where and when to adapt Flow packets contain advice about how to control congestion and may be stored at the network node

12. Best Effort MPEG Delivery Partial Packet Discard - discard packets on buffer overflow Static Priority Discard - two level priority scheme Frame Level Discard – queue a datagram iff its corresponding frame can be entirely queued Group of Picture Level Discard – if I-frame is dropped, drop corresponding P & B frames.

13. Performance Analysis Performance metric : fraction of received data not discarded D i,k – fraction of discarded data T i,k – performance of model i at a source rate of k Mbps i = { P, S, F, G }

14. Performance Analysis Performance metric : signal-to-noise ratio

15. Conclusion Active networking is consistent with, and even suggested by the E2E argument Active networks outperform the end-to-end solutions ?