Research On Physical Layer Authentication Of Industrial IOT Wireless Communications

The Industrial Internet of Things(IIoT)can effectively improve industrial production effi-ciency.Therefore,the IIoT has attracted tremendous attention from researchers and engineers in recent years.The realization of the IIoT has stringent requirements for communication systems,including:low latency,high reliability and high security.Traditional industrial scenarios use wired cables to connect various industrial equipment,which can meet the strict IIoT communi-cation requirements.However,the application of wireless technology is the trend of the IIoT.It can bring many benefits:low cost,high long-term stability,and it is suitable for communication devices that are difficult to connect with wired cables.While wireless communication technol-ogy brings these benefits to the IIoT system,it also introduces some challenges.Shadows and fading of wireless channels affect the reliability of communication systems,and communication failures may increase system delay.Besides,the broadcast characteristic of wireless channels makes the IIoT more vulnerable to security attacks.In order to solve these challenges,we use non-coherent Massive Single-Input Multiple-Output(Massive SIMO)technology based on en-ergy modulation to build a high-reliability and low-latency IIoT system.Physical layer authen-tication schemes are proposed for this system,including physical layer message authentication(active form)and physical layer identity authentication(passive form).Through theoretical analysis and simulation experiments,we verify the performance of the proposed schemes.First,we use the non-coherent Massive SIMO technology based on energy modulation to build a communication system for the IIoT,which can achieve low-latency and high-reliability communication.In the uplink,multiple sensor devices communicate with the controller in a time division multiple access manner.We provide a communication process flow to ensure se-cure authentication,including identity authentication and message authentication.For the IIoT system,the impact of different security attack models and the corresponding security strategies are introduced.Spoofing attacks,tampering attacks,and replay attacks are more harmful to the IIoT system.Authentication mechanism is a key solution to the security problems.We define the evaluation metrics of the considered system from the aspects of communication reliability and authentication security.It is theoretically analyzed that the application of non-coherent technology can reduce communication delay,and Massive SIMO technology can improve sys-tem reliability.The correctness of the theoretical analysis is verified by simulation experiments.And the influence of different system parameters on the communication performance is studied.By selecting appropriate parameters,the symbol error rate(SER)of the communication system can be reduced to less than 10^-8.Secondly,we propose a physical layer message authentication method based on the em-bedded message authentication code(MAC).It can be applied to the IIoT system to achieve reliable authentication with low latency.Closed-form expressions of the SER of the message signal and the tag signal(obtained by MAC modulation)are derived from the embedding model.The performance of the traditional uniform embedding scheme is verified through simulation experiments.The simulation results show that there is a tag SER error floor close to 10%in the traditional uniform embedding scheme,which cannot meet the reliability and security requirements of the IIoT.This finding prompts us to propose a new embedding design,i.e.,”Message-based tag embedding”.In this paper,the embedding design problem is modeled as an optimization problem and is further transformed into a convex optimization problem.The optimal embedding design is obtained by solving the convex optimization problem.Next,sim-ulation experiments are performed to verify the correctness of the theoretical derivation,and the performance of the optimal embedding design is given.The SER trade-off curve for message signals and tag signals also characterizes the trade-off between system communication reliabil-ity and security.It provides a reference for actual system designers.Finally,we study the physical layer identity authentication based on the channel state in-formation(CSI)through an actual industrial data set.This data set collects channel frequency response in two scenarios:static and mobile industrial communications.It also provides the corresponding device location.The controller needs to collect the CSI of legitimate devices in advance and establish a corresponding database.For static scenarios,the controller uses the channel response difference as a basis to determine whether the message comes from a legiti-mate device.We propose an improved choice of the channel difference.It uses the average CSI value of the legitimate device as the standard for calculating the channel difference.The test results show that the detection probability is increased from 97.92%to 100%when the false alarm probability is 0.However,in the mobile scenario,the above authentication method is not applicable.We propose a solution based on the channel difference between adjacent moments for mobile scenarios.The test results show that the detection probability reaches 98.61%when the false alarm probability is 0.