Downlink Resource Allocation And Block Error Rate Performance Analysis In URLLC

As the rapid development of wireless communication,URLLC is becoming a typical sce-nario in next generation,which is required to support ultra high reliability and low latency simultaneously,the reliability should no less than 1-10^-5.Finite blocklength（FBL）based short packet communication is used for URLLC to reduce latency,which has becoming an important enabler.In the mean time,NOMA is a key technology in next generation.How-ever,the combination of finite blocklength communication and NOMA is still in its infancy.To investigate the benefit of the combination,this thesis compare the performance of NOMA and OMA in finite blocklength communication from the perspective of resource allocation and BLER performance analysis.The main work of the thesis is summarized as follows:1.The theory of finite code length transmission and the basic principle of power domain NOMA are introduced.Aiming at the resource allocation problem of URLLC downlink,the solution algorithm is given and the simulation analysis is carried out.Finally,the theoretical analysis and simulation verification of the block error rate performance are carried out.2.For the URLLC downlink,with the block error rate of the worst link as the optimiza-tion target,a resource allocation problem based on minimum and maximum fairness is pro-posed,and this problem is transformed into an optimization problem of minimizing total power.Derivative-based iterative binary search algorithm is proposed to solve the problem.The simulation results show that the decoding error probability of the system can decrease by increasing the blocklength or power.Through an effective resource allocation strategy,NOMA can achieve a lower block error rate than OMA.3.For the URLLC downlink of NOMA and OMA,the block error rate performance is analyzed respectively,the analytical expression of each user’s block error rate in the two-user system is deduced,and the approximate expression when the number of transmitted bits is small is deduced.Different from the traditional analysis method of order statistics,this thesis analyzes the block error rate performance of each user by synthesizing all possible decoding orders,and obtains the mathematical expectation of block error rate.Finally,the correctness of the theoretical deduction is verified by simulation.The results show that for a delay-constrained downlink,the NOMA transmission mechanism based on dynamic decoding order proposed in this thesis can achieve a lower block error rate than OMA.