Research On Cooperative Nonorthohonal Multiple Access And Physical Layer Security Technology For Power Line Communications

Power line communication(PLC)is one of the most effective communication methods for information exchange in the power Internet of Things.But,at the same time,the surge of the number of communication devices and the explosive growth of data transmission requirements in the power Internet of Things will also have higher requirements for the throughput,rate,connection density and security of PLC.Non-orthogonal multiple access(NOMA)technology allows multiple users to share time-frequency resources and has high spectrum efficiency,which is expected to meet multi-dimensional performance requirements such as the high throughput,large connection,and high rate.Cooperative relaying technology can effectively resist the channel fading by exploiting its space diversity,thus enhancing the communication reliability as well as extending the network coverage.The performance of communication system can be significanlty improved by combining the two technologies and taking the advantages of both.In addition,physical layer security(PLS)technology uses the randomness of channels to secure communication,which has great potential in enhancing the security of system.However,due to the uniqueness of PLC and its complexity of communication environment,the related research is still in its infancy,and there are many problems to be solved,such as the spectral efficiency and reliability of cooperative NOMA PLC system to be improved and the lack of discussion on the complex and severe PLS problems.Therefore,based on the above reasons,this dissertation is committed to studying the cooperative NOMA and PLS technologies of PLC in the context of the application of power Internet of Things,and focuses on dealing with the above problems to be solved to realize the efficient,reliable and secure communication.This dissertation carries out the research on PLC cooperative NOMA transmission mechanism with higher spectrum efficiency,reliability and security.The main contents are as follows:Firstly,to deal with the problems of low channel utilization and spectrum efficiency in the traditional PLC cooperative NOMA transmission scheme,an incremental relaying based PLC adaptive cooperative NOMA transmission strategy is proposed.By introducing the feedback of nearby user signal reception quality,this strategy allows the source node to send the next data signal to the nearby user when the relay node forwards the distant user signal,or let the nearby user decode its desired signal again by listening to the signal forwarded by the relay,thus making full use of idle links during the second data transimission phase.Considering the PLC log-normal channel fading and BernoulliGaussian noise,the system throughput performance of this strategy in delay-limited applications is analyzed.Also,the system throughput analysis of both the traditional cooperative NOMA and the orthogonal multiple access transmission schemes are selected as the benchmarks.The simulation experiments verify the superiority of developed strategy,and reveal the impacts of key parameters,such as the impulse noise and relay location.Secondly,the space diversity is further used to enhance the system reliability,and the multi-relay cooperative NOMA PLC system and the opportunistic relay selection mechanism are studied.A PLC cooperative NOMA opportunistic relay selection scheme based on the maximum weighted harmonic mean is then proposed in order to solve the problems of the poor adaptability to time-varying channels caused by fixed user sequencing and power allocation and to improve the system energy efficiency,thus alleviating the electromagnetic radiation of PLC.Specifically,under both the decode and forward and amplify and forward relaying protocols,according to the quality of service(Qo S)requirement of each user,the user sequencing and user signal’s power allocation are dynamically adjusted by using the instantaneous channel state information(CSI)of PLC links,and then the optimal relay is selected by taking the criterion of minimizing system power consumption.The outage and throughput performance of system and their closed-form expressions are further given under the PLC log-normal fading channels disturbed by impulse noise.The final simulation results verify the superiority of the proposed scheme.Then,the PLS problem of PLC cooperative NOMA transmission is studied to secure communication.In the external eavesdropping scenario with considering the affects caused by the imperfect successive interference cancellation(SIC),a PLC cooperative NOMA secure transmission strategy assisted by artificial noise(AN)and secrecy performance optimization method are proposed.By allowing the relay to transmit the AN signals while forwarding the confidential NOMA signals to deteriorate the eavesdropping quality.To comprehensively evaluate the reliability and security of system,the effective secrecy throughput performance of each user and its asymptotic performance are analyzed.In addition,in order to improve the secrecy performance,the secrecy outage probability(SOP)of each confidential signal is analyzed,and then with the SOP and power constraints,the redundancy rate of confidential signals and the power allocation between the confidential signals and AN signal are optimized to maximize the system secrecy sum rate.Then,the model-free deep reinforcement learning(DRL)method is applied to solve this non-convex optimization problem in real time,and then a hierarchical DRL algorithm is proposed by leveraging the hierarchical strategy idea and reasonably designing the DRL state and immediate reward.Finally,simulations verify the superiority of the strategies propoed in this chapter,and evaluate the effects of SIC imperfections and impulse noise.Finally,the PLS problem of cooperative NOMA PLC system in the untrusted multirelay eavesdropping scenario,where the security threat is more serious,is further studied.And with considering the influence of the imperfect CSI,a DRL based PLC untrusted relaying cooperative NOMA is proposed.In order to deteriorate the untrusted relays’ eavesdropping quality,the source node consumes partial power to generate and send the AN signals when transmitting the confidential NOMA signals.Then,to maximize the the system secrecy sum rate,a problem of robust secure relay selection and power allocation is constructed under the constraints of the users’ Qo S,power and channel uncertainty constraints introduced by imperfect CSI.In order to solve this complex joint optimization problem in real time,avoid relying on the system model,and deal with the network scale changes in real communication system,a hierarchical DRL algorithm based on quantized CSI is proposed.This algorithm decomposes the joint optimization problem using the hierarchical strategy idea,and then exploits the deep Q-network to learn the optimal strategy for each sub-problem in decomposed action spaces;the DRL state and action are designed based on the quantized CSI’s interval index of communication links,making the algorithm’s the structure independent of the network scale.The final simulation verifies the superiority of this DRL robust secure transmission scheme.