Cascade self‐tuning control architecture for QoS‐aware MAC in WSN

Abstract

Wireless sensor network (WSN) has proposed requirements more than just data collection and transmission. Different types of traffics call for different quality of service (QoS) demands. However, for the nature of uncertainty of wireless link and unpredictability of traffic transmission, it is very hard to distribute resources to different traffics precisely ahead of time. Based on the cascade self-tuning control, this study proposes the cascade self-tuning architecture for FD-MAC (CSFD-MAC) architecture. Taking the advantage of feedback theory, CSFD-MAC decouples contention window (CW) adaption and active time (AT) adaption into two separated cascading control loops. CW controller aims for the absolute node-to-node delay to a fixed value and AT controller guarantees the relative delay ratio. The software simulations illustrate the validity of controller design and the hardware experiments demonstrate the ability of CSFD-MAC in the absolute delay control as well as the proportional delay differentiation. More important, CSFD-MAC which takes AT as a controlled variable, not only has a much more superior performance in the metrics of delay and throughput compared with other QoS-aware MAC, but also has the advantage of efficient power management.