Error pattern analysis for data transmission of wireless sensors on rotating industrial structures

Abstract

Wireless sensors capable of sensing, processing, and wireless communication have been adopted for monitoring purposes in a variety of contexts, many of which feature challenging radio propagation characteristics. Fast rotating structures are commonly found in mechanical, transportation and energy systems, and the challenges of using wireless sensors on such structures have not been adequately addressed. For wireless sensors on rotating structures, it has been found there is an eminent dependency of packet error rates on rotation speeds, bursty bit errors, periodic variation in received signal strengths, and dominance of multipath effects. Previously, a reliable data transmission approach was developed to recover transmission errors, and it was found that transmission errors have mostly occurred in certain high-error regions. The objective of this study is to utilize the transmission approach to infer the transmission error pattern of such regions. The paper presents a sliding window algorithm for estimating the error region width, and a transmission simulator for studying dependencies among error burst statistics, error region distribution and widths, and accuracy of the rotation speeds. It is concluded that i) the number, width, and distribution of error regions can be inferred from normalized error burst distance distributions, ii) the error burst distance distributions depend sensitively on accuracy of the rotation speed, and iii) the error in speed can be inferred from the error burst distance distributions. The conclusions directly guide future work on transmission error modeling, on-line error pattern inference, and error-avoidant transmission methods for radios on rotating structures.