Thursday, August 25, 2016

SIGCOMM 2016 - Session 8 - Wireless - Paper 1: "Inter-Technology Backscatter: Towards Internet Connectivity for Implanted Devices".

Presenter: Vikram Iyer
Authors: Vikram Iyery, Vamsi Tallay, Bryce Kelloggy, Shyamnath Gollakota and Joshua R. Smith

Summary:
Smart medical devices such as smart contact lens, implantable devices and etc. requires are crucial part in pursuing treatment and cure of chronicle diseases, and have high requirement on nework connectivity.

The driving question of this paper is: can these implanted devices communicate directly with mobile devices such as smartphones, watches and tablets? The key challenge is that these devices cannot use conventional radios such as WiFi, Bluetooth and ZigBee because these radios consumes too much power while these devices are extremely power constrained. 

The solution proposed in this paper is to recycle radio frequency signals from external devices, aka, backscattering. Though many papers propose solutions using this technology, they tend to require extra hardware support. In this paper,  the authors introduce interscatter, a novel backscatter communication system that works using only commodity devices by transforming transmissions from one technology to the other, on the air.  For instance, interscatter can create 2–11 Mbps Wi-Fi standards-compliant signals by backscattering Bluetooth transmissions.

The basic idea of interscatter involves two steps. The first step is to create single-tone transmission using Bluetooth radio. The intuition is that Bluetooth uses two frequencies to encode the zero and one data bits. By transmitting a stream of constant ones or zeros, a single frequency tone can be created. To cope with the Bluetooth data whitening,  reverse engineering can be used by setting different original payload bits. This is because the whitening sequence in Bluetooth protocol uses a pseudo random bit generator that fixes the channel number as the seed. 

The second step of interscatter is to  generate 802.11b signals by backscattering the single-tone from the Bluetooth device. Previous works on passive WiFi can achieve this goal, but also creates a mirror copy outside the ISM band. Interscatter proposes the first single sideband backscatter architecture that produces a frequency shift on only one side of the single tone Bluetooth transmission. The intuition in this architecture is to generate orthogonal signals so that their mirror copy signal can be cancelled out.

A prototype backscatter hardware on an FPGA platform is implemented. Evaluation shows that interscatter can generate 2–11 MbpsWi-Fi signals from Bluetooth transmissions. And several proof-of-concept applications such as a contact lens form-factor antenna and an implantable neural recording interface antenna are also implemented and evaluated.


Q&A:
Q1: Is it possible to charge these implant device wirelessly whiling performing inter-backscattering?
A: Wireless charging is not the focus of our paper. But it will be an interesting topic to study the feasibility of wireless charge and backscatter simultaneously.

Q2. What is the next step of backscatter?
A: We are hoping more involvement of the community in developing products using this technology. In terms of inter-backscattering, our current applications are proof-of-concept, and we are working on design product-ready devices.

Q3. How small is the contact-lens prototype? 
A: The proof-of-concept prototype is smaller than WiFi/Bluetooth chips. (See Sec. 5.1)

Q4. Is this technology going to be open-sourced?
A: It is already licensed to a start-up.