Self-Management in Chaotic Wireless Deployments

1. Self-Management in Chaotic Wireless Deployments Aditya Akella, Glenn Judd, Srini Seshan, Peter Steenkiste Presented by: Farhana Ashraf Adapted from Mobicomm 2005 presentation by Glenn Judd

2. Wireless Proliferation • Sharp increase in deployment • Airports, malls, coffee shops, homes… • 4.5 million APs sold in 3rd quarter of 2004!

3. Change in Wireless Landscape UNPLANNED and UNMANAGED PLANNED and MANAGED CHAOTIC NETWORK!!!

4. Outline • Quantify characteristics of chaotic wireless deployment • Impact on end-user performance • Initial work on mitigating negative effects • Conclusion

5. Characterizing Current Deployments

6. 50 m 1 1 2 1. AP Degree [Place Lab] City #AP Max.degree A B C

7. Degree Distribution [Place Lab]

8. Most users don’t change default channel Channel selection must be automated 2. Unmanaged APs [WifiMaps] Channel % AP

9. 3. AP Management Support [Wardrive] Vendor % AP • Major vendors dominate • Incentive to reduce “vendor self interference”

10. Outline • Quantify characteristics of chaotic wireless deployment • Impact on end-user performance • Initial work on mitigating negative effects • Conclusion

11. Impact on Performance [Globosim Trace-driven simulation on Wardrive Data set] • 20 APs in topology • Each AP has D clients • Each client runs HTTP/FTP workloads • Vary stretch ‘s’  scaling for inter-AP distances

12. Impact on HTTP Performance 3 clients per AP. 2 clients run FTP sessions. All others run HTTP.300 seconds 5s sleep time Degradation 20s sleep time Max interference No interference

13. Optimal Channel Allocation vs.Optimal Channel Allocation + Tx Power Control Channel Only Channel + Tx Power Control

14. require: mediumUtilization <= 1 txPower determines range dclient, txPower determines rate dclient dmin Impact of Joint Transmit Power and Rate Control Objective: given <load, txPower, dclient> determine dmin APs

15. Impact of Transmit Power Control • Minimum distance decreases dramatically with transmit power • High AP densities and loads requires transmit power < 0 dBm • Highest densities require very low power  can’t use 11Mbps!

16. Outline • Quantify characteristics of wireless deployment • Impact on end-user performance • Initial work on mitigating negative effects • Conclusion

17. Power and Rate Selection Algorithms • Fixed-Power Rate Selection • Auto Rate Fallback: ARF • Based on probe • Estimated Rate Fallback: ERF • Based on SNR DATA If ACK fails, Decrease rate If success, Increase rate B A ACK

18. Power and Rate Selection Algorithms • Adaptive-Power Rate Selection • Sender reduces power as long as rate is not reduced • Extension to Fixed-Power Rate selection • Power Auto Rate Fallback: PARF • Power Estimated Rate Fallback: PERF

19. Lab Interference Test Topology Results

20. Conclusion • Significant densities of APs in many metro areas • Many APs not managed • High densities could seriously affect performance • Static channel allocation alone does not solve the problem • Transmit power along with adaptive rate control effective at reducing impact