Portrait of Ranveer Chandra

Ranveer Chandra

Chief Scientist
Microsoft Azure Global


My research focuses on different aspects of mobile computing, in particular wireless communications and energy efficiency. Currently I am working on three different projects:

    • White space networking in the TV bands and other portions of the spectrum (KNOWS Project page)
    • Low latency wireless for gaming and other real-time applications
    • Energy-efficient computing for smartphones and tablets (Battery Project Page)

Video Links: An overview of my research is available in this MSR Luminaries Video. Some other video links are here: on white spaces, on improving battery life, and on code offload.

In the past I have worked on wireless management, mesh networks, enterprise network management, and different wireless architectures. More details about some of the projects are below.

White Space Networking (Cognitive Radios, SDRs, …)http://research.microsoft.com/knows

We started the KNOWS project at Microsoft Research in early 2006, and since then I have led the technical aspects of the project through three versions. In the first phase, we looked at the MAC protocol for white spaces in a mesh setting. In the second phase, we proposed WhiteFi, a system for providing Wi-Fi like connectivity for white spaces. In the third phase, we have now built and deployed a white space network in Microsoft campus. More details on this project can be found here.

Energy-Efficient Computing

I am working on ways to reduce the energy consumption of PCs and mobile phones. Somniloquy (NSDI 2009) provides a way to let PCs go to sleep mode without losing network connectivity. Cell2Notify is a notification architecture for mobile phones that allows devices to turn on Wi-Fi only when needed, thereby increasing their battery lifetime. This is an active project, and expect to see more information soon.

Network Management (eXpose, DAIR, Sherlock, …)

With the increasing complexity of enterprise networks, there is a pressing need for efficient approaches to detect, localize, diagnose and repair faults in the network. I have worked on several aspects of network management as part of a bigger project, called NetHealth. Sherlock uses dependency analysis to localize faults that impact end-users, DAIR helps IT administrators locate regions of poor WLAN coverage and performance, and WiFiProfiler enables users to collaborately diagnose Wi-Fi problems.

“Whats Going On? Extracting Communication Rules in Edge Networks”, Srikanth Kandula, Ranveer Chandra, Dina Katabi. Proceedings of ACM SIGCOMM, Seattle, August 17-22, 2008

Mobile Systems

Smartphones are becoming extremely popular. While they provide a new and unique user experience, several key problems need to solve to fully realize their potential. First, RF spectrum is limited, and a proliferation of smartphones has led to significantly degraded performance. This is likely to become worse in the future. Second, battery technology is not evolving at the same pace as applications, and the energy-hungry applications on a smartphone quickly drain its battery. With these constraints in mind, I have been working on technologies and applications to make better use to the RF spectrum while consuming lesser energy.

  • “MAUI: Making Smartphones Last Longer with Code Offload”, Eduardo Cuervo, Aruna Balasubramanian, Dae-ki Cho, Alec Wolman, Stefan Saroiu, Ranveer Chandra, Paramvir Bahl, Proceedings of ACM MobiSys 2010, June 15-18, 2010.
  • “Virtual Compass: Relative Positioning to Sense Mobile Social Interactions”, Nilanjan Banerjee, Sharad Agarwal, Paramvir Bahl, Ranveer Chandra, Alec Wolman, Mark Corner, Proceedings of Pervasive, May 17-20, 2010.
  • “Wi-Fi Neighborcast: Enabling Communication Among Nearby Clients”, Ranveer Chandra, Jitendra Padhye, Lenin Ravindrananth. Proceedings of ACM HotMobile, Napa Valley, February 25-26, 2008.
  • “BeaconStuffing: Wi-Fi Without Associations”, Ranveer Chandra, Jitendra Padhye, Lenin Ravindrananth, Alec Wolman. Proceedings of IEEE HotMobile, Tucson, February 26-27, 2007.
  • “Wireless Wakeups Revisited: Energy Management for VoIP Over Wi-Fi Smartphones”, Yuvraj Agarwal, Ranveer Chandra, Alec Wolman, Paramvir Bahl, Kevin Chin, Rajesh Gupta. Proceedings of ACM/USENIX MobiSys, Puerto Rico, June 11-14, 2007

Mesh Networks

I have worked on networking issues in layer 2 and layer 3 (both on unicast and multicast) in wireless mesh networks. The details of this project are available on our group’s Mesh Networking page.

Rethinking the Design of Wireless Systems

Wi-Fi is designed as a convenient replacement for the wire, and its design mostly attempts to mimic the behavior of the Ethernet. For example, a Wi-Fi card needs to associate to an Access Point (AP) before it can send data packets, which is similar to the action to plugging the Ethernet cable in the wall. Furthermore, Wi-Fi restricts a card to associate to at most one AP. In my opinion, the wireless medium has a number of powerful properties and wireless networks should be treated differently from traditional wired networks. I have built some systems that try to break away from these constraints in Wi-Fi networks as part of VirtualWiFi and WiFiAds projects.