Delay monitoring for wireless sensor networks: An architecture using air sniffers

Wireless sensor networks have been used for many delay-sensitive and safety–critical applications, e.g., emergency response and plant automation. For such applications, delay measurement inside the sensor networks is important for real-time monitoring and control of the networked system, and abnormal delay detection. In this paper, we propose a measurement architecture using distributed air sniffers. This approach provides convenient delay measurement, and requires no clock synchronization or instrumentation at the sensor nodes. Since using sniffers incurs additional deployment cost, we investigate two aspects to reduce deployment cost: (1) using inexpensive mote-class sniffers and (2) carefully placing the sniffers to minimize the number of sniffers that are needed. Specifically, we experimentally quantify the capability and fidelity of mote-class sniffers for delay measurement, and show that they provide satisfactory monitoring performance. We further formulate and solve a sniffer placement problem that minimizes the number of sniffers while taking account of the workload constraints of the sniffers, and show that the number of sniffers under our sniffer placement algorithms is only a small fraction of the number of sensor nodes in the network. Last, we demonstrate the effectiveness of our architecture for abnormal delay detection using experiments in a testbed.