Research On Resource Management And User Scheduling For PD-NOMA Based Cellular Networks

In recent years,novel services such as autonomous driving,ultra HD video and smart home have emerged.As the current network is difficult to meet the requirements,the next-generation mobile communication technology for the future has become a global research and development hotspot.The development of each generation of mobile communication systems has always been accompanied by the emergence of various multiple access technologies.Traditional Orthogonal Multiple Access(OMA)usually allocates separate time and frequency radio resources to each user.It has been unable to meet the requirements of higher spectrum efficiency and larger number of connections in future wireless networks.Non-Orthogonal Multiple Access(NOMA)allows multiple users to share the same resource,greatly increasing system capacity and user connections,and reducing device power consumption.NOMA has become one of the key technologies for future communication networks.This dissertation focuses on downlink resource allocation and user scheduling for the power domain NOMA based cellular networks.The technology still has the following shortcomings: Since all users in the NOMA occupy the same time-frequency resources,and signals of different users need to be distinguished in the power domain,power allocation in the NOMA is crucial.In addition,the power allocation,user clustering and precoding have mutual influence and constraints.The computational complexity of the joint optimization algorithm is too high,and the current low complexity algorithm has poor performance,which is difficult to meet the requirements of future networks;Successive Interference Cancellation(SIC)is the key technology for signal separation of different users in NOMA.However,due to the error rate of SIC receiver and channel estimation error,SIC may have errors,and subsequent errors may affect subsequent signal separation.However,there are currently few studies on SIC errors;In future wireless communication networks,power consumption is an important factor to be considered in network design.However,the research on multi-cell energy-saving technology based on NOMA technology is still not mature,and the theory of resource management and user scheduling under dynamic NOMA network is less.In view of the above problems in the current NOMA technology,this dissertation decouples the user clustering and power allocation problem by using channel correlation,avoiding the high complexity brought by joint optimization.Since users with high channel correlation are assigned to the same cluster,a zero-forcing algorithm can be used for precoding design to minimize inter-cluster interference.Then,a low-complexity power allocation algorithm is designed by convex optimization,and the influence of SIC error on the system is also improved from the perspective of power allocation.In the heterogeneous networks,the downlink NOMA system model is firstly innovatively established,and the capacity of each state of the near and far users is analyzed from the statistical point of view.On this basis,the coverage of the base station in the multi-cell network is variable and the random walk of the user is studied.The algorithm is proposed to optimize the system power consumption and handoff.In summary,the proposed algorithm achieves better performance indicators under low complexity and promotes the application of NOMA in heterogeneous networks,which has some theoretical and practical value.