Research On Throughput Optimization Mechanism Of Backscatter Link In Passive Sensing System

As an important bridge between the physical and information world,the Internet of Things(IoT)is becoming more and more digital,ubiquitous and intelligent.Passive sensing has attracted worldwide attention due to its functions of passive,maintenance-free,perceptive computing,communication and networking characteristics,which provide the realization possibility for the diversification of the application of IoT,and the sharp increase of data volume will be inevitable.With the development of Computational Radio Frequency Identification(CRFID)platform and multi-energy source environment Backscatter technology,passive sensing system is gradually being used on a large scale.Among them,the problem of throughput optimization becomes increasingly important.In large-scale IoT applications or scenarios with high real-time requirements,when a large amount of sensor data collected by tag needs to be transmitted or when the tag is in a state of rapid movement,the sensor data information needs to be transmitted to reader as soon as possible.Nowadays,the passive sensing system is throughput constrained when it applied in large scale scenarios or burst data transmission,in view of the bottleneck ofhigh detection channel overhead,low effective utilization rate of time slot and poor robustness of parallel decoding algorithm,this paper conduct research on throughput optimization of Backscatter link in passive sensing system from the aspects of protocol optimization and parallel decoding,respectively.Firstly,this paper optimizes the protocol design,introduces the related indicators of signal cluster in IQ domain from physical layer,utilize more fine-grained Cluster Gather Degree(CGD)and Ambiguous Symbol Probability(ASP)indicators of signal cluster in IQ domain to monitor the current channel state,so as to realize rate adaption on physical layer,which makes the data transmission rate to match the channel conditions and realizes the maximization of throughput under different channel conditions.Secondly,protocol design on MAC layer combines the probability and duration of collision and idle slot,the number of clusters in current collision slot is also introduced to optimize the Q value selection algorithm that directly determines the number of time slots,so that the Q value can be adjusted to match the number of tags as soon as possible to improve the utilization rate of time slot.In addition,on the aspect of parallel decoding,based on the clock synchronization and Multi—Modulus Algorithm(MMA),this paper minimize the environmental dynamics caused by inter-symbol interference and error from the source,which makes the signal cluster in IQ domain show the structural characteristics.Then first order absolute difference is made to signal sampling point in IQ domain,using the number of signal flow clusters in IQ domain after absolute difference is onemore than the number of tags under big probability,in view of the expected situation under big probability and the unexpected situation under small probability,this paper combined with probability by difference characteristics to realize parallel decoding,further multiplied improve the throughput of Backscatter link and expand the application scenario of passive sensing system.Finally,this paper takes the typical CRFID application platform,Wireless Identification and Sensing platform(WISP)as the experiment platform,the Universal Software Radio Peripheral(USRP)was regarded as the reader and monitor,so as to monitor the physical information in real time.On the basis of this,the proposed design in our paper is implemented and the evaluated.The final experimental results show that the throughput optimization of proposed algorithm increased significantly,and the robustness promotion effect under dynamic environment and multi-tag scenarios is remarkable.