An Empirical Study of Industrial Real-Time Wireless Mesh Network in Field Deployments

Real-time wireless sensor and actuator networks (WSANs) have experienced widespread adoption across many industries due to their great advantages of enhanced mobility, easier deployment as well as reduced configuration and maintenance costs. To compensate for the dynamic and unpredictable nature of wireless networks, extensive research efforts have been devoted on algorithm design for efficient network resource management to both meet robustness requirements and optimize the overall network for latency, power, and throughput. Very limited results, however have been reported for the network performance in real field deployments. This paper presents an in-depth exploration of a WirelessHART mesh network deployed in a separations research facility. To understand how the network performs under different scenarios it is restarted and data traces are collected at three points in time with different experiment settings. We examine the impact of initial neighbor selection and observe the network manager adapting the network. The mesh network performance is reported in terms of its primary functions, which are to report process measurements and maintain the network operations.