Can WebRTC QoS Work? A DSCP Measurement Study

DiffServ was designed to implement service provider quality of service (QoS) policies, where ingress and egress routers change the DiffServ Code Point (DSCP) in the IP header. However, nowadays, applications are beginning to directly set the DSCP themselves, in the hope that this will yield a more appropriate service for their respective video, audio and data streams. WebRTC is a prime example of such an application. As a first step towards understanding whether "WebRTC QoS works", we measured, for both IPv4 and IPv6, what happens to DSCP values along Internet paths. Our study is based on end-to-end measurements from 160 IPv4 and 65 IPv6 geographically spread controlled probe clients to 34 IPv4 and 18 IPv6 servers respectively. Clearly, when the DSCP value is changed, the net result may not be what the application desired. We find that this happens often, and conclude with recommendations on how to improve WebRTC and other applications using the DSCP.

[1]  Ming Zhang,et al.  An untold story of middleboxes in cellular networks , 2011, SIGCOMM.

[2]  Ramesh Govindan,et al.  Investigating Transparent Web Proxies in Cellular Networks , 2015, PAM.

[3]  Michael Welzl,et al.  How to say that you're special: Can we use bits in the IPv4 header? , 2016, ANRW '16.

[4]  Gorry Fairhurst,et al.  Tracking transport-layer evolution with PATHspider , 2017, ANRW.

[5]  Mark Allman,et al.  A middlebox-cooperative TCP for a non end-to-end internet , 2014, SIGCOMM.

[6]  David L. Black,et al.  Definition of the Differentiated Services Field (DS Field) in the IPv4 and IPv6 Headers , 1998, RFC.

[7]  Cullen Jennings,et al.  DSCP Packet Markings for WebRTC QoS , 2016 .

[8]  Olivier Bonaventure,et al.  Revealing middlebox interference with tracebox , 2013, Internet Measurement Conference.

[9]  Brian Trammell,et al.  Observing Internet Path Transparency to Support Protocol Engineering , 2015 .

[10]  Sally Floyd,et al.  On inferring TCP behavior , 2001, SIGCOMM.

[11]  Gorry Fairhurst,et al.  Exploring DSCP modification pathologies in mobile edge networks , 2017, 2017 Network Traffic Measurement and Analysis Conference (TMA).

[12]  Stein Gjessing,et al.  fling: A Flexible Ping for Middlebox Measurements , 2017, 2017 29th International Teletraffic Congress (ITC 29).

[13]  Gorry Fairhurst,et al.  PATHspider: A tool for active measurement of path transparency , 2016, ANRW '16.

[14]  David Black,et al.  Differentiated Services (Diffserv) and Real-Time Communication , 2015, RFC.

[15]  Thomas Dreibholz,et al.  NorNet Core - A multi-homed research testbed , 2014, Comput. Networks.

[16]  Sally Floyd,et al.  Measuring the evolution of transport protocols in the internet , 2005, CCRV.

[17]  Vern Paxson,et al.  Here Be Web Proxies , 2014, PAM.

[18]  Larry L. Peterson,et al.  The design principles of PlanetLab , 2006, OPSR.

[19]  Florian Wohlfart,et al.  Analysis and topology-based traversal of cascaded large scale NATs , 2013, HotMiddlebox '13.

[20]  Mark Handley,et al.  Is it still possible to extend TCP? , 2011, IMC '11.

[21]  Harald Alvestrand,et al.  Overview: Real-Time Protocols for Browser-Based Applications , 2021, RFC.

[22]  Sally Floyd,et al.  Measuring interactions between transport protocols and middleboxes , 2004, IMC '04.

[23]  Fred Baker,et al.  Configuration Guidelines for DiffServ Service Classes , 2006, RFC.