USTC UBIoT's paper Multiprotocol Backscatter with Commodity Radios for Personal IoT Sensors was collected by IEEE/ACM ToN in October 2022.
IEEE/ACM ToN (Transactions on Networking) is a leading international journal in the field of communication/computer networking. It is also a CCF-A journal recommended by the Chinese Computer Society.
Here is a brief introduction to the paper:
Thesis title: Multiprotocol Backscatter with Commodity Radios for Personal IoT Sensors
Authors: Yuan Longzhi, WANG Qiwei, Zhao Jia, Gong Wei
Summary of the paper content:
Backscatter communication is a novel communication technology. Its nodes do not need to generate radio frequency signals alone, but rely on existing radio frequency signals to carry information, so as to reduce power consumption and cost. At present, a large number of related work using various commercial signals has emerged, which provides the basis for the large-scale deployment of backscatter communication. However, these works usually assume that the backscatter node knows the signal type, which is difficult to guarantee in real application scenarios. In addition, the backscatter modulation of existing compatible commercial equipment usually requires the deployment of two receivers to obtain the information before and after the node modulation, and the combination of the two to demodulate the node data. To solve the first problem, this paper proposes to use the rectifier circuit on the backscatter node to extract the envelope information of the radio frequency signal, and identify the degree of correlation between it and several reference waveforms of the protocol identification target. This method can achieve more than 97% accuracy for the most common radio signals of 2.4GHz, including 802.11b, 802.11n, Bluetooth, ZigBee, etc. To solve the second problem, we asked the sender to generate data packets containing duplicate symbols, and the backscatter node only uses the second and later symbols to carry node data. The receiver can obtain data packet information before and after node modulation at the same time, so a single receiver can demodulate node data.
In order to extract baseband information of radio frequency signal, a diode as the core rectifier circuit is built. In order to obtain high quality and high amplitude baseband information, we set the bandwidth of the rectifier circuit to 40MHz.
We use a low power ADC to collect the baseband envelope waveform through the rectifier circuit, and calculate its cross-correlation value with the reference waveform of several identified objects on FPGA. When the correlation between the envelope waveform and a reference signal is much greater than that of other reference signals, we set the identification result to the protocol type of this reference signal.
After recognition, the backscatter node modulates the RF signal according to the recognition result to transmit its data. The experimental results show that the node can achieve more than 97% recognition accuracy, and successfully use the recognition results to achieve the targeted modulation of different carriers, up to 250kbps data rate.
