A Passive Sensing Parallel Transmission Technique Based On Spatial And Temporal Distribution Of Signals

Backscatter provides the benefits of energy harvesting and low power communication,making it attractive to a broad class of applications.These applications,such as warehouse inventories,object tracking,and industry surveillance,involve large volumes of data carried by a large number of deployed tags.Therefore,how to fully utilize the channels and enhance network throughput,turns to be a crucial problem in the relevant area.One potential solution to provide high throughput is to parallelize multiple backscatter transmissions.The existing parallel transmission approaches,however,achieve very low throughput in practice.One major reason is that decoding the highly complicated collision signal requires much higher signal quality.However,in real-world application scenarios,like libraries,retail stores,and warehouse,most of tags are deployed distant from the reader,shielded by racks,or even buried under other objects,which lead to particularly low signal strength.To solve this problem,we propose Signal Scope,a reliable parallel decoding method which achieves high decoding rate even under low SNR scenarios.The contribution of this paper can be summarized as follows:1.Building the Markov-based signal moving modelIn this paper,we show through experiments that the existing parallel decoding methods exhibit particularly low decoding rate in practical scenarios.The reason behind is they cannot work under low SNR scenarios.Specifically,under low SNR,signal clusters at the IQ domain will overlap with each other,making the signal transitions buried under the noises,and meanwhile results in particularly low decoding rate.Therefore,before decoding,we should first identify the state of each received signal sample.We find through experiments that changing a perspective to observe the moving of the signal helps us to extract the underlying state transition sequence of the signal.Based on this observation,we propose a Markov-based model(named Flow model)to trace the moving pattern of the signal.The Flow model provides important clue for the subsequent signal state tracing component.2.Reliable signal state tracing based on the Flow model.The first challenge we meet is to extract the state of the signal(i.e.,estimating the number of signal states and find the locations of the clusters which represent different signal states).In this paper,we propose a reliable method to extract the signal states based on signal's spatiotemporal distribution.This method achieves high accuracy even under low SNR scenarios.Meanwhile,we further propose a state identification method,which identify the state of each signal sample based on the Flow model.Finally,an error correction scheme is designed to further guarantee the reliability of parallel decoding.