Channel Estimation For Battery-free Backscatter Communication Systems

As the third revolution of world information industries,future Internet of Things(IoT)aims to interconnect everything,changing the pattern of traditional communication systems.It is also one of the five emerging strategic industries,and will provide efficient and convenient services in many fields,such as intelligent transportation,public safety,environmental conservation,industrial monitoring and information collection.Meanwhile,the fifth generation(5G)wireless networks will increase transmission rates tenfold or even hundredfold with extensive access and diverse IoT data processing.In turn,IoT will offer efficient network configuration for 5G to satisfy various terminal requirements.Clearly,5G will boost the popularity of IoT.However,there exist a series of challenges for IoT.The high cost of radio frequency(RF)components and limited battery life are two well-known challenges for wireless sensors in IoT.Recently,battery-free backscatter,a green communication technology,attracts much attention.It enables sensors or tags to harvest energy from ambient wireless signals and then communicate with each other by backscattering the signals.The battery-free backscatter technology can free sensors from battery constraints and can also low their costs due to no requirement of RF components.It is worth noting that there are fundamental differences between passive backscatter and traditional Radio Frequency Identification(RFID)including carrier signal sources,received signals and variable channel parameters.Therefore,existing communication theories cannot be directly applied in battery-free backscatter systems.Accurate channel state information can improve the detection performance and is a key factor in transceiver design and security enhancement.Channel estimation for battery-free backscatter systems is an open problem,which motivates our current work.The channel estimation problem for battery-free backscatter is a challenge due to the following two reasons:(1)There are hidden variables in the systems including signals from the RF source and the tag.The channel parameters vary with different states of the tag;(2)Few or limited training symbols can be transmitted from batteryless tags due to power constraint.This paper studies two system models for battery-free backscatter and proposes two channel estimators.(1)In the first scenario,signals of RF source remain unknown to the receiver.We propose a blind channel estimator based on the expectation maximization(EM)algorithm and define the intermediate variables to acquire the modulus values of channel parameters.We also obtain the ranges of the initial values of the suggested estimator and derive the modified Bayesian Cramer-Rao bounds.Simulation results are then provided to corroborate our theoretical studies.(2)In the second scenario,the communication protocol of RF source is public and a few signals,such as training symbols,are known to the receiver.We design a communication protocol for the reader and the tag,and propose a semi-blind channel estimator based on Least Square(LS)and EM algorithms to acquire combined channel parameters.We also obtain the uplink and downlink channel parameters between the reader and the tag through maximum likelihood(ML)estimation with superimposed pilots from the reader.In addition,we derive the estimation error lower bounds of all channel estimates.Finally,simulation results illustrate the performance of our estimators and confirm the validity of the theoretical analysis.