Final Report Summary - CHAOSNETS (Building Scalable, Secure, and Reliable "Chaotic" Wireless Networks)
As a result of their unplanned, license-free nature, Wi-Fi networks have proliferated in recent years, giving users unprecedented improvements in wireless access to the Internet. But being "chaotic," i.e. completely unplanned, they have grown to be victims of their own success: when eager users set up too many wireless access points in a densely populated area, the resulting noise and interference hurt everyone’s throughput and connectivity. The CHAOSNETS project is a ground-up redesign of chaotic wireless networks, with new architectural contributions focusing on what information the physical layer should pass up to higher layers. We have developed a new physical layer interface called SoftAoA that passes angle-of-arrival (AoA) information from the physical layer up to higher layers. Using this expanded physical layer interface, we have first innovated in error control coding and receiver-based rate adaptation methods to improve wireless capacity in the vagaries of the ``grey zone'' of marginal coverage. Second, we have transformed the areas of wireless device localization, environmental sensing, and wireless security with our new ideas based on the profiling of incoming packets' AoA at nearby access points. Finally, we have investigated how chaotically-deployed access points that are willing to loosly cooperate can mitigate the interference they experience from other networks not under the same administrative control, and manage the interference they cause to those networks. A derivative ERC Proof-of-Concept project (SmartTap) has commercialized CHAOSNETS' second focus area and resulted in multiple European, International, and US patents pending. CHAOSNETS' results have been disseminated in world-leading computer networking venues including the ACM SIGCOMM and MobiCom Conferences, and the USENIX NSDI Symposium, garnering multiple Best Paper awards, as well as a runner-up Best Doctoral Dissertation Award from the British Computer Society. CHAOSNETS has thus demonstrated more scalable, secure, and reliable chaotic wireless networks that play an even more prominent role in our lives.