Meeting control performance over a wireless mesh network

Good control performance requires that the overall control loop execute 4 to 10 times faster than the process time constant plus deadtime. When calculating the overall control performance the amount of time required to read input signals, execute control, and drive outputs must be taken into consideration. These times are referred to as latency. The variation in input, output, and control execution is referred to as jitter. Too much latency or jitter can severely degrade overall control performance. Managing the latency and jitter in the input and output operations is challenging in wireless networks. Wireless signal strength varies in time; wireless nodes come and go arbitrarily; data is routed dynamically on different paths. Oversampling burns batteries out. Missing or delay of the process data by the network degrades control performance. In this paper we present an approach to meet the control performance requirements using a wireless mesh network. We utilize a time synchronized mesh network with short time slots. We argue that device and network operation must be synchronized. We discuss the technical challenges in this approach and show how these challenges can be overcome. We base this approach on the emerging WirelessHARTtrade standard.