Key Technologies Of Resource Management In Narrowband Internet Of Thing And Dual Connectivity Systems

In recent years,with the rapid growth of mobile Internet,artificial intelligence,Internet of Thing and intelligent mobile phone,the interconnection demand of all things becomes more and more strong.Every people and object will be networked in future programmable world.The ultra high-speed Internet connection becomes more and more urgent for intelligent mobile phone especially for indoor mobile users,and the network capacity and traffic demand will increase by thousands of times in the next decade.Evolution of current cellular networks is needed to meet the requirements of future network services.The deployment of low power wide area networks can achieve interconnection of all things,and the deployment of small cells can improve the system capacity and traffic.Finally,the NarrowBand Internet of Thing(NB-IoT)and Dual Connectivity(DC)are proposed to satisfy the requirements of wide area IoT and ultra high-speed Internet connection respectively,and complement each other.However,the related issues such as wide coverage,strong mobility and high security need to be well explored.Firstly,we introduce the advantages and key technologies of NB-IoT and DC,which include the key techniques of physical layer and media access control layer in NB-IoT,user plane,control plane architecture and power control in DC.Secondary,we proposed two-dimension uplink link adaptation algorithm based on repetition character,proposed the joint congestion physical layer security capacity model.The power control algorithm and small cell downlink flow control optimization is studied deeply to improve the system overall performance.The main contents and innovations of this paper are as follows:(1)We propose a novel MCS-dominated link adaptation algorithm for NB-IoT according to its properties such as low power consumption,large coverage,huge capacity and repetition.The proposed link adaptation algorithm includes inner loop link adaptation and outer loop link adaptation.MCS and repetition are determined effectively with the cooperation of inner and outer loop link adaptation.The performance of our proposed link adaptation algorithm is verified by link level simulation.About 14%transmit time and resource units can be saved compared to repetition-dominated link adaptation algorithm,and above 46%transmit time and resource units can be saved compared to the traditional link adaptation algorithm.(2)We design a cooperative amplify and forward jamming strategy of relay nodes for physical layer security.The simulation results show that the signal quality can be significantly improved by diversity gain of relay node,after making full use of wireless channel time-varying,multi-path and fading character.Security capacity can be greatly enhanced and ensures that the information can not be overheared by the eavesdropper.For low power consumption and limited memory NB-IoT nodes,physical security is an effective complement for complex key encryption algorithm.The combination of physical security and encryption from upper layer can ensure the security performance,save energy consumption and prolong life time.(3)We proposed the interference management strategy of channel and power control based on the non-cooperative game theory in DC.The uplink power and channel allocation is modeled as a non-cooperative game process after considering channel fading and layer interference.The utility function is set as the small cell total capacity.The proposed algorithm not only can effectively reduce same layer interference and improve system throughput,but also lower complexity.As for the downlink power control problem,optimal power allocation shceme is designed with the aim of minimizing the total power consumption.The proposed optimization problem uses transmit-powers of macro and small cells as variables,and it is difficult to solve due to natural non-convexity.Despite of this,we transform it into an equivalent traffic scheduling problem by exploiting the feature of DC.The transformed traffic scheduling problem is proved to be a Monotonic Optimization(MO)problem which can be optimally solved by the Polyblock Approximation(PA)method.Numerical results are presented to demonstrate the performance and advantages of our proposed traffic offloading scheme with security consideration.(4)As for non-ideal backhaul DC scenario with more than 5ms delay,we proposed a pricing based flow control algorithm by using the input of cell average flow statistics,small cell target delay and user buffer status report.The purpose is to enable the user to minimize total data cost.The simulation results demonstrate that too small target delay will result in obvious decrease of user throughput,slightly larger target delay such as 10-20ms is recommended;flow control period will also influence user throughput,small flow control period shch as 5-10ms is recommended.Theoretical analysis and simulation results show that appropriate flow control parameters and interference aware resource allocation strategy can greatly improve throughput of small cell,which close to the throughput of carrier aggregation,and achieve the lowest total data cost.Finally,the resource management researches of NB-IoT and DC are summarized,and certain future development and improvement directions are pointed out.