1 / 23

Improving TCP Performance over Ad-hoc Network

Improving TCP Performance over Ad-hoc Network. Course Project for EE228A --- Fall 2000 (Professor Jean Walrand). Xuanming Dong, Duke Lee, and Jin Wang. 11/28/2000. Outline. TCP basics and congestion control Ad-hoc network basics Impact of ad-hoc network on TCP performance

scott
Télécharger la présentation

Improving TCP Performance over Ad-hoc Network

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Improving TCP Performanceover Ad-hoc Network Course Project for EE228A --- Fall 2000 (Professor Jean Walrand) Xuanming Dong, Duke Lee, and Jin Wang 11/28/2000

  2. Outline • TCP basics and congestion control • Ad-hoc network basics • Impact of ad-hoc network on TCP performance • Explicit Route Notification • Simulation and Results • Summary

  3. Basic function of TCP • Basic Data Transfer: use IP services; • Reliability: sequence number; positive ACK; retransmission and time out; checksum; • Flow Control: • Congestion Control: • Multiplexing: addresses or ports within each host; • Connections: • Precedence and Security:

  4. TCP: Congestion Control • Variables: • Congestion Window (W) & Slow Start threshold size (ssthresh) • Indication of congestion: • Timeout and DupACK • Congestion Control : • Slow Start + Congestion Avoidance • Fast Recovery (Fast retransmission + Congestion Avoidance)

  5. Ad-hoc Networks • Autonomous mobile nodes are free to move around arbitrarily • network topology may change randomly and rapidly at unpredictable times • Without (necessarily) using a pre-existing infrastructure • Routes between nodes may potentially contain multiple hops

  6. Infrastructure-oriented Wireless Network • Have fixed and wired gateways • The bridges for these networks are known as base stations • A mobile node within these networks connects to the nearest base station • As the mobile node travels out of range of one base station and into range of another, a “handoff” occurs • Usually only the last link is wireless

  7. How does mobility affect TCP • High bit error rate • cause packets(TCP data segment or ACKs) corrupted and then congestion control invoked • Route recomputation • Discovering a new route may take significantly longer than RTO and then cause retransmission and invoke congestion control • Network may be temporarily partitioned • Cause packets dropped and invoke congestion control

  8. Route Re-computation Figure 1: Route Re-Computation

  9. Temporary Network Partition Figure 2: Temporary Network Partition

  10. TCP sender times out. Resumes sending mobility causes link breakage, resulting in route failure TCP sender times out. Backs off timer. Route is repaired No throughput No throughput despite route repair Larger route repair delays especially harmful TCP data and acks en route discarded Why Does Throughput Degrade?

  11. Explicit Route Notification(ERN) Router-based feedback • Need to modify TCP Source & Router to receive/send feedback • Need mechanisms for information exchange between layers • Inform TCP of route failure by explicit message • Let TCP know when route is repaired By Explicit notification • Reduces repeated TCP timeouts and backoff

  12. TCP FSM

  13. Explicit Route Notification(ERN) Router-based feedback • Need to modify TCP Source & Router to receive/send feedback • Need mechanisms for information exchange between layers • Inform TCP of route failure by explicit message • Let TCP know when route is repaired By Explicit notification • Reduces repeated TCP timeouts and backoff

  14. Support from Router • Maintain a packet cache for every routing entry in the routing table. Once the router finds that the routing entry for a packet becomes invalid recently, it caches the packet and creates an ERFN (Explicit Route Failure Notification) packet. • Associate a timer with every packet in the cache. If the timer expires, then drop the packet from the cache. • If the invalid routing entry is removed from the routing table, drop all packets in its cache and cancel the associated timers. • If the invalid routing entry becomes valid again, forward all the cached packets and create an ERRN (Explicit Route Recovery Notification) packet for each cached packet.

  15. Support from Router

  16. Support from TCP • Context Switch • The Frozen State • Stop sending further packets. • Stop invoking congestion control. • All TCP timers are frozen, except the new frozen timeout timer. • But TCP source may accept ACK packets and advance the acknowledgment window.

  17. Support from TCP

  18. Simulation Results • Source Packet Trace • Source Congestion Window trace • TCP Throughput

  19. Simulation Results

  20. Simulation Results

  21. Simulation Results

  22. Simulation Results

  23. Summary By Xuanming Dong

More Related