DeWOZ: Rethinking the Schmidl-Cox Algorithm for Detecting Wi-Fi Out of ZigBee Interference

Abstract

This letter presents detect Wi-Fi signals out of ZigBee (DeWOZ), a clean-slate technique to discern Wi-Fi packets drowning in ZigBee interference. Unlike existing solutions that are restricted to the cases where Wi-Fi signal strength is much higher than ZigBee, Wi-Fi signals only interfere with ZigBee payload, or beforehand clear Wi-Fi transmissions are done to obtain transmitter-specific information, DeWOZ aims at achieving the detection of relatively weak Wi-Fi signal hidden in both ZigBee header and payload portions only based on a single corrupted signal. DeWOZ re-explores the Wi-Fi-oriented Schmidl-Cox algorithm for corrupted signal synchronization followed by two key operations, autocorrelation anomaly detection and adaptive discrete Fourier transform, to sniff Wi-Fi out of ZigBee header and payload, respectively. The feasibility of DeWOZ is validated via practical experiment.