Estimation Theory For New Battery-free Backscatter Communication Systems

The core idea of the Internet of Things(Io T)is the Internet of everything.The development of this technology breaks the limitation that only certain devices can access the network.The rapid development of 5G technologies not only bring faster network,but also provide reliable support for future Io T.However,with the rapidly increasing number of connected devices,large-scale communications will face a series of challenges,including energy sources and hardware costs for sensors,large-scale access methods,composite channel estimation and complex signal detection.As a potential technology to solve the energy and cost problems of Io T devices,the new backscatter communication technology proposed by academia has become one of the key technologies for future Io T and also one research hotspots in the field of mobile Internet.Traditional backscatter systems usually require specific radio frequency source power,often resulting in short communication distance.In recent years,new battery-free backscatter technologies,including ambient backscatter,have been put forward.These new battery-free backscatter technologies can utilize the RF energy in the existing communication system to achieve green communications,and have important application values and prospects for future Io T.Channel estimation is an important basic theory in wireless physical layer,which has significant influence on the overall system performance.However,the channel estimation theory for battery-free backscatter systems is not yet mature and needs further investigation.Due to the differences between passive backscatter systems and traditional backscatter systems in carrier signals and channel parameters,existing estimation methods cannot be directly applied to battery-free backscatter systems.Meanwhile most current studies assume simple system models with flat fading channels,which is different from the actual application scenario.Thus this thesis aims to investigate design of channel estimators for new battery-free backscatter communication systems.To begin with,this thesis introduces new battery-free backscatter communication systems and the corresponding communication processes.The fading characteristics of the wireless channels are analyzed and new channel models are established.Then existing channel estimators are discussed.Next,two new battery-free backscatter systems models for the following two different scenarios are constructed and the corresponding channel estimators are designed.(1)In the first urban scenario,frequency-selective channels are more suitable to modeling the practical situations due to the complexity of the surrounding environment.Since the channel parameters and the received signals are different when the tag is in various states,i.e.,reflecting and non-reflecting states,an iterative estimator based on the Least Square(LS)algorithm is proposed together with the communication protocol and pilot structure.We also derive the CRLBs for each channel estimates.Simulation results are then provided to corroborate our theoretical studies.(2)In the second high-speed mobile scenario,unmanned aerial vehicle and other devices are usually used as RF sources,and thus the battery-free backscatter systems can be established over doubly selective channels.Assuming that the tags can take advantage of the energy from existing Orthogonal Frequency Division Multiplexing(OFDM)signals,we propose a new model-driven channel estimator by designing the communication protocol and the comb pilot structure.This suggested estimator combines the traditional LS algorithm with the neural network to improve estimator performance.We also analyze the complexity of the new estimator.It is found that the complexity of our new estimator is similar to that of the traditional Linear Minimum Mean Square Error(LMMSE)algorithm.Finally,simulation results show that the proposed new estimator outperforms traditional LMMSE algorithm.