Measuring the Performance of User Traffic in Home Wireless Networks

This paper studies how home wireless performance characteristics affect the performance of user traffic in real homes. Previous studies have focused either on wireless metrics exclusively, without connection to the performance of user traffic; or on the performance of the home network at higher layers. In contrast, we deploy a passive measurement tool on commodity access points to correlate wireless performance metrics with TCP performance of user traffic. We implement our measurement tool, deploy it on commodity routers in 66 homes for one month, and study the relationship between wireless metrics and TCP performance of user traffic. We find that, most of the time, TCP flows from devices in the home achieve only a small fraction of available access link throughput; as the throughput of user traffic approaches the access link throughput, the characteristics of the home wireless network more directly affect performance. We also find that the 5 GHz band offers users better performance better than the 2.4 GHz band, and although the performance of devices varies within the same home, many homes do not have multiple devices sending high traffic volumes, implying that certain types of wireless contention may be uncommon in practice.

[1]  Stefan Savage,et al.  Automating cross-layer diagnosis of enterprise wireless networks , 2007, SIGCOMM '07.

[2]  Suman Banerjee,et al.  Observing home wireless experience through WiFi APs , 2013, MobiCom.

[3]  Paramvir Bahl,et al.  Architecture and techniques for diagnosing faults in IEEE 802.11 infrastructure networks , 2004, MobiCom '04.

[4]  Suman Banerjee,et al.  Diagnosing Wireless Packet Losses in 802.11: Separating Collision from Weak Signal , 2008, IEEE INFOCOM 2008 - The 27th Conference on Computer Communications.

[5]  Konstantina Papagiannaki,et al.  Experimental Characterization of Home Wireless Networks and Design Implications , 2006, Proceedings IEEE INFOCOM 2006. 25TH IEEE International Conference on Computer Communications.

[6]  David Kotz,et al.  Analysis of a Campus-Wide Wireless Network , 2002, MobiCom '02.

[7]  Nick Feamster,et al.  BISmark: A Testbed for Deploying Measurements and Applications in Broadband Access Networks , 2014, USENIX ATC.

[8]  Konstantina Papagiannaki,et al.  PIE in the Sky: Online Passive Interference Estimation for Enterprise WLANs , 2011, NSDI.

[9]  Nick Feamster,et al.  Broadband internet performance: a view from the gateway , 2011, SIGCOMM.

[10]  Matti Siekkinen,et al.  Performance Limitations of ADSL Users: A Case Study , 2007, PAM.

[11]  Justin Manweiler,et al.  RxIP: Monitoring the health of home wireless networks , 2012, 2012 Proceedings IEEE INFOCOM.

[12]  Srinivasan Seshan,et al.  Can user-level probing detect and diagnose common home-WLAN pathologies , 2012, CCRV.

[13]  Suman Banerjee,et al.  Airshark: detecting non-WiFi RF devices using commodity WiFi hardware , 2011, IMC '11.

[14]  Ratul Mahajan,et al.  Analyzing the MAC-level behavior of wireless networks in the wild , 2006, SIGCOMM.

[15]  Yin Zhang,et al.  On the characteristics and origins of internet flow rates , 2002, SIGCOMM '02.