Abstract
Factory automation systems based on the IEEE 802.11 Wi-Fi standard may benefit from its Multi-Rate Support (MRS) feature, which allows for dynamically selecting the most suitable transmission rate for the targeted application context. The MRS is implemented by means of rate adaptation algorithms (RAAs), which has already demonstrated to be effective to improve both timeliness and reliability, which are typically strict requirements of industrial real-time communication systems. Indeed, some of such algorithms have been specifically conceived for reliable real-time communications. However, the computational complexity of such algorithms has not been effectively investigated yet. In this paper, we address such an issue, particularly focusing on the execution times of some specific rate adaptation algorithms, as well as on their impact on the automation tasks. In this respect, after a formal description of the algorithms, we present the outcomes of an extensive experimental session, which includes practical measurements and realistic simulations. The obtained results are encouraging, since the measured execution times indicate that rate adaptation algorithms can be profitably adopted by industrial automation systems, allowing for improving their reliability and timeliness without impacting on the overall performance.
Highlights
Modern factory automation systems are based on distributed architectures in which communication networks are exploited to connect industrial controllers and field devices, like sensors and actuators, located throughout a plant, possibly at considerable distances [1]
In order to achieve deeper insights about the various components of a data exchange cycle, we evaluated the time taken by the communication sequence of Figure 2, that in the following we refer to as round trip time (RTT), and the consequent impact of the rate adaptation algorithms on it
We addressed the impact of rate adaptation algorithms on the task execution times of distributed automation systems, based on Wi-Fi
Summary
Modern factory automation systems are based on distributed architectures in which communication networks are exploited to connect industrial controllers and field devices, like sensors and actuators, located throughout a plant, possibly at considerable distances [1]. A typical cause of such variability is represented by a failure in delivering a packet, which necessarily requires a retransmission: in this case, the actual task communication time results at least doubled with respect to the time necessary to a single transmission attempt This problem is even more evident when wireless networks are used, since they are inherently error prone [5,6]. Several RAAs have been designed over the years and some of them were conceived for industrial applications [12,13] These latter ones have demonstrated their effectiveness, in that they are able to rapidly adapt to the SNR behavior, and revealed to be able to ensure low communication times, along with less variability and limited packet loss ratios. This paper moves from the above considerations and focuses on the computational aspects of rate adaptation algorithms, as well as of protocol stacks, adopted for industrial real–time applications, as will be described in the following
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
