Logo image
A Flexible Retransmission Policy for Industrial Wireless Sensor Actuator Networks
Conference proceeding

A Flexible Retransmission Policy for Industrial Wireless Sensor Actuator Networks

Ryan Brummet, Dolvara Gunatilaka, Dhruv Vyas, Octav Chipara and Chenyang Lu
2018 IEEE International Conference on Industrial Internet (ICII), pp.79-88
10/2018
DOI: 10.1109/ICII.2018.00017

View Online

Abstract

Real-time and reliable communication is essential for industrial wireless sensor-actuator networks. To this end, researchers have proposed a wide range of transmission scheduling techniques. However, these methods usually employ a link-centric policy which allocates a fixed number of retransmissions for each link of a flow. The lack of flexibility of this approach is problematic because failures do not occur uniformly across links and link quality changes over time. In this paper, we propose a flow-centric policy to flexibly and dynamically reallocate retransmissions among the links of a multi-hop flow at runtime. This contribution is complemented by a method for determining the number of retransmissions necessary to achieve a user-specified reliability level under two failures models that capture the common wireless properties of industrial environments. We demonstrate the effectiveness of flow centric policies using empirical evaluations and trace-driven simulations. Testbed experiments indicate a flow-centric policy can provide higher reliability than a link-centric policy because of its flexibility. Trace-driven experiments compare link-centric and flow-centric policies under the two reliability models. Results indicate that when the two approaches are configured to achieve the same reliability level, a flow-centric approach increases the median real-time capacity by as much as 1.42 times and reduces the end-to-end response times by as much as 2.63 times.
Industrial Job shop scheduling Real Time Real-time systems Reliability Reliable Communication Runtime Schedules Wireless Wireless communication Wireless sensor networks

Details

Metrics

Logo image