RFID Based Liquid Sensing

Recently,with the rapid development of RFID,more IOT applications have en-tered our life,including warehouse management,positioning,activity sensing,etc.As the key technology of IOT field,the task of RFID has changed from recognition to wireless sensing.Yet the study of sensing mechanism has become a research hotspot.Leveraging the mechanism of backscatter,RFID can detect the state of environment,e.g.,liquid,by the influence of environmental interference.In this paper,we focus on the RFID based liquid sensing.We explore the physical meaning of RFID signal feature and the influence of liquid.Because of the special physical characteristics of water molecule,Liquid is able to affect RFID signal by absorption,reflection and refraction.Our experiments show that the signal feature,e.g.,RSSI,phase,will be influenced by these phenomena in RFID system with liquid.We will describe them by constructing their physical and mathematical model,which is the technical bedding of our certain workFirst,we propose an RFID-based approach for monitoring water level through an RFID tag array attached on a container.Due to the absorption of liquid,the energy of RFID backscatter signal will suffer attenuation,which also depend on the distance be-tween liquid and passive tag.Based on this,we capture the RSSI vector continuously,then use a signal match model to detect the water level.In particular,we put forward a method to adjust profile dynamically.We have done some performance evaluation,our experiment shows that out algorithm can be used to accurately predict the water level over 85%of the time and the average height deviation is around 0.2cm.Based on the approach,we implement our real system named RF-iCare,an RFID-based in-fusion detecting system.RF-iCare is aim to detect the amount of water in the infusion and to alert.Considering the real world and the demand of users,for more convenient purpose,we design the hardware and software.RF-iCare has now been deployed in several hospitals.Our experiment shows that RF-iCare can accurately alarm when the water level is low over 90%of the time.Second,we propose an RFID-based approach for monitoring drip speed.We just attach RFID tag array to the Murphy's dropper.With the reflection and refraction ef-fect of liquid,RFID tags' phase value will be changed when the water drop down.We design a system frame to detect the frequency of drip Leveraging the periodic change.To address the phase periodicity and time slot asynchronization,we first address pre-process the data,then filter the signal to eliminate noise,finally extract frequency us-ing FFT-based method.Especially,we introduce low-pass filter and DWT-based filter method to remove interference signal.We implement our system and evaluate the per-formance,our result shows that it can achieve an average frequency deviation of 3.1 drops per minute.The main contribution in this paper are as follows:1.We summarize the physical meaning of RFID signal feature,and construct the physical and mathematical model to describe the influence of liquid,including absorption,reflection and refraction.2.Leveraging the absorption of liquid,we propose an RFID-based approach for monitoring water level.In particular,we put forward a method to adjust pro-file dynamically for robustness.We propose RF-iCare,an RFID-based infusion detecting system to detect the amount of water in the tagged container and to alert.We design the hardware and software for more convenient purpose.Then RF-iCare now has been deployed in several hospitals.3.Leveraging the reflection and refraction of liquid,we propose an RFID-based approach for monitoring drip speed.Especially,we introduce low-pass filter and DWT-based filter method to remove interference signal for promoting robust-ness.