Towards Interference-aware ZigBee Transmissions in Heterogeneous Wireless Networks

Cross-technology interference (CTI) from diverse wireless networks such as ZigBee, Bluetooth, and Wi-Fi has become a severe problem in the 2.4 GHz Industrial Scientific and Medical (ISM) band. Especially, low power and lossy networks are vulnerable to the signal interferences from other aggressive wireless networks when they perform low power operations to conserve the energy consumption. This paper presents CoSense, which accurately detects ZigBee signals with a reliable signal correlation scheme in the presence of the CTI. The key concept of CoSense is to reduce false wake-ups of low power listening (LPL) by identifying the pre-defined ZigBee signatures. Our scheme is robust in the coexistence environment of diverse wireless technologies since the signal correlation works well in bad wireless channel conditions. It achieves standard compliance and transparency without any hardware and firmware changes. We have implemented CoSense on the Universal Software Radio Peripheral (USRP) platform to verify its feasibility. The experimental exploration reveals that CoSense significantly reduces the false-positive and false-negative rate under typical setting and the additional overhead is negligible. The results show that our scheme saves much energy by up to 63% in dynamic network interference scenarios where low-power ZigBee transmissions are overwhelmed by strong Wi-Fi signal interferences.