1. An End-to-end Approach to Increase TCP Throughput Over Ad-hoc Networks Sarah Sharafkandi and Naceur Malouch

2. 2 Introduction TCP is designed for wired networks  Congestion control : window-based With IEEE 802.11 PHY & MAC, TCP over Ad-hoc has a low performance:  congestion control and not “collision” control: TCP react to buffer overflow   " bursty " traffic  inherent reverse traffic Objective: Improve TCP throughput without modifying PHY, MAC and NET layers.

3. 3 When collision causes DATA loss? By hidden nodes: packets sent by D collide with A’s packets at node B preventing B from decoding A’s packets. By repetitive retries due to “ordinary” collisions: it happens when   C*  rare event By buffer overflow : due to increased waiting times  not considered in this work

4. 4 State of the art Distributed Link RED and Adaptive pacing [Fu et al. INFOCOM’2003]  If the average number of retransmission retry > min_thresh : early drop of packets increase the backoff period   Improvement: 10%-30% for the chain topology Increasing retry limit and optimum packet size [Jiang et al. DISCEX’O3]  Increasing the retry limit reduces oscillations in the instantaneous thpt  Increasing the packet size increases the thpt till some thresh Improving TCP throughput using Delayed ack method [Altman et al. MADNET’03]  delayed ack factor = 2, 3

5. 5 An end-to-end approach to “collision control” ?!

6. 6 Simulation Scenario NS2 network simulator Chain topology The source and destination at both ends of the chain AODV as a routing protocol Some modifications to the source code of NS2:  delayed ack > 2  monitoring without file traces  token bucket: packet version

7. 7 TCP Sends the packets in “burst” Two experiments to show the effect of “burstiness”  Simulation with TCP using RFC3465  Simulation with CBR traffic

8. 8 Simulation with TCP using RFC3465  The “burstiness” of RFC3465 results in throughput reduction despite the gain in the window growth

9. 9 Simulation with CBR traffic: Results Best result is when there is packet spacing  “burstiness” is minimum i CBR traffics with rate r/i, i = 1, 2, 3, 4.

10. 10 New approach Bursty data traffic over Ad-hoc networks results to performance reduction Shaping :  Controls the rate of releasing packets to the network  No more aggressive traffic  Plus delayed ack  approaches the optimal channel reuse

11. 11 Throughput of TCP with shaper and delayed ack Shaper increases the TCP throughput by 53%-120%

12. 12 Shaper and Delayed ack Shaper allow delayed ack mechanism to bypass the limit of d=3 

13. 13 Optimum rate There is always an optimum rate for the shaper in which TCP has the best performance

14. 14 TCP throughput as a function of Number of hops Optimum rate decreases when number of hops increases

15. 15 Impact of bucket size A data can pass through the shaper only if it can get a token from token buffer. We can use it to test again the effect of burstiness

16. 16 Tokens Again allowing “burstiness” results to throughput reduction

17. 17 Effectiveness of Shaping in presence of CBR Traffic Network scenario : same source/destination for UDP traffic  UDP share all the ad-hoc routers with TCP Compute the gain while increasing the rate of UDP:

18. 18 Conclusion TCP throughput drops significantly because of: link contention caused by hidden terminal problem An "aggressive“ TCP sender causes an increased contention at the MAC layer Implementing a shaper at the sender improves TCP throughput by controlling the aggression of TCP data traffic Delayed ack mechanism plus the shaper → increase spatial channel reuse

19. 19 Future work An adaptive algorithm for finding the optimum rate  difficulties: convergence and stability  Related work: [ElRakabawy et al. MobiHoc’2005] same idea: end-to-end solution BUT :  change TCP protocol for the multihop wireless ad-hoc  based on the esimation of the 4-hop transmission delay  Our approach :