No abstract

We have designed and implemented a wearable ZigBee-based guidance system in which a main controller collects ultrasound and GPS signals from sensors attached to subcontrollers and provides appropriate directions to a blind person. The signals from the different sources need to be handled separately, which reduces the effective sampling rate, and may impact the quality of the guidance provided. To address this, we propose a new scheduling scheme that interleaves the two signals without collision, based on the on-line estimation of signal handling times. Experimental results show that this scheme increases a sampling rate by up to 71% compared with the simple scheme 1 .

No abstract

No abstract