1. WiFox: Scaling WiFi Performance for Large Audience Environments Arpit Gupta, Jeongki Min and Injong Rhee NC State University

2. Interesting ?? Faster WiFi !! Manageable solution !! WiFox: AP-only S/W solution WiFox: AP-only S/W solution

3. Large Audience Environments (LAEs)

4. Large Audience Environments Any location with a large WiFi user population WBA projected a growth rate of 350% per year for such WiFi deployments Various successful deployment models such as Boingo-Google, Mobily-Aruba etc. already exist Source: WBA(Wireless Broadband Alliance) Source: Wireless Broadband Alliance (WBA)

5. What about User Experience ? As number of active clients increases, user experience diminishes significantly

6. Problem Anatomy

7. Well Known Factors Contention and collision – Increases with growing competition Rate Diversity – Slower STA slows down all other STAs – Various fairness realizations like WFQ, TBR etc. Random Losses and TCP performance – TCP treats packet losses as congestion signals – Usage of TCP ECN and proxy servers isolate wired and wireless networks Traffic Asymmetry

8. Downlink Traffic dominates for 90% of data traces

9. Traffic Asymmetry Majority of data traffic is Web based Downlink Traffic dominates compared to uplink Majority of data packets are for HTTP based web activities

10. Traffic Asymmetry In scope of our problem it is: – Downlink/Uplink Asymmetry – Channel Access Asymmetry Implications: – Packets spend more time at AP’s TxQ – Frequent packet drops

11. Wireless Channel Uplink Traffic Downlink Traffic Channel access for uplink traffic is more

12. Performance Bottleneck TxQ saturates as associated clients increase Associates Users

13. Goodput Performance Traffic Asymmetry causes TxQ saturation resulting in poor goodput performance

14. Possible Solutions

15. Wireless Channel Uplink Traffic Downlink Traffic Equal Channel Access for Uplink/Downlink

16. Wireless Channel Uplink Traffic Downlink Traffic Statically Assign Higher Priority to Downlink Traffic

17. Wireless Channel Uplink Traffic Downlink Traffic Dynamically Assign Higher Priority to Downlink Traffic

18. Our Solution

19. Priority Control Packet A ACK DIFS DIFS H Channel Access N Slots Busy medium Wins Contention STA C STA B STA A Smaller IFS ClassCWminCWmaxAIFSTXOP Limit AP15164 STAs510N/A

20. Linear Scaling Priority Model Priority Level Linear relationship between Goodput and Priority Level

21. Adaptive Prioritization 100 ms DHDHDDDHDD Time High Priority Default Priority Decision Points Priority Level 3

22. Evaluation

23. Test Bed 2600 Sqft Area with multiple APs, 45 STAs Netgear 802.11 b/g wireless cards with Atheros chipsets and MADWIFI drivers Latency emulation using DummyNet Modified SURGE for web traffic generation Requests inter-arrival closely follows the ones observed for SIGCOMM traces Uplink UDP traffic using Iperf to emulate Background Traffic

24. Performance Downlink: N/W Goodput WiFox W/O WiFox Significant improvement in Network’s Downlink Goodput

25. Performance Experiment involves sending 25 requests and observe response for 4 minutes duration Request Serving rate is 4 times better than NPC Experiment involves sending 25 requests and observe response for 4 minutes duration Request Serving rate is 4 times better than NPC

26. Robustness Performance in presence of Multiple Aps ?? WiFox w/o WiFox

27. Robustness w/o WiFox WiFox Unfair Distribution Fairness Realization ??

28. Performance: TxQ Dynamics WiFox w/o WiFox

29. Conclusion WiFox Delivers: – Deployment Ready Solution – Enhanced user experience with 400-700% Downlink Goodput improvements 40-60% faster response time Open Problems: – Characterizing asymmetry problem for 802.11n – Support for real time applications like chats etc. – QoS

30. 30 Merci !!

31. Multi AP Scenario DDDDDD DDDDDDD AP 1 ( Priority Level 4) AP 2 ( Priority Level 3) DDDDDDD DDDDD AP 1 ( Priority Level 3) AP 2 ( Priority Level 5) time 100 ms

32. Performance: Insight Enables AP to switch to HIGH priority state under heavy load Avoids TxQ saturation Significant reduction in ReTx rate compared to stock WiFi implementation

33. Robustness: Uplink Traffic Scenarios where few users indulge in heavy uplink activities like video uploading, cloud synchronization etc.

34. Observations Source: Rodrig et al.