Research On Massive Unsourced Random Access Technology Based On Polar Code

With the development of global informatization,the scale of the Internet of Things industry continues to increase,opening a new era of Internet of Everything.As one of the main application scenarios of the Internet of Things,massive machine-type communication has attracted extensive attention in the industry.Among them,the characteristics of massive connection,low energy consumption,short data packets and sporadic communication for massive machine-type communication demonstrate the fundamental differences between traditional cellular networks.The conventional multiple access and coding scheme is no longer applicable.Aiming at the new requirements of massive random access of Internet of Things,this paper conducts preliminary research and exploration on unsourced massive random access technology,and designs a massive access coding and communication schemes with low complexity and energy-efficient.Firstly,the existing solutions of massive random access are studied,the T-fold ALOHA protocol is im-proved,and a random access coding scheme based on T-fold IRSA is given.The simulation analyzes the relationship between the channel load and the unknown probability q of the data packet and the effect of dif-ferent T values on the system performance.And it is concluded that the relationship between the normalized load threshold and the T value is not proportional.Only when the T value is small,the normalized load threshold is larger and the system throughput is higher,and when the maximum number of packet repetitions is 2,2≤T≤4,the system performance is optimal.On the other hand,polar codes are not only proved to achieve the capacity of any binary memoryless symmetric channel when the code length tends to infinity,but also have shown to outperform low-density parity-check codes in the case of short code length.Therefore,this paper designs a polar code scheme based on T-fold IRSA.Every message is split into two parts.The first part is the preamble sequence,which is used to select the slot and improve the anti-interference ability be-tween users.Due to the natural sparsity of the vector composed of active users in the massive random access scenario,the receiver uses the compressed sensing decoder to recover.The experiment simulates the effect of different parameters N_p、k_pand K_bon the performance of the compressed sensing decoder.The results show that the preamble sequence can be almost completely recovered by selecting an appropriate parame-ters.The second part is the payload,which uses polar code for channel coding to improve anti-interference ability of the channel.At the transmitter,the frozen bit selection scheme is given.At the receiver,a joint successive cancellation list decoding algorithm is studied to recover the multi-user information sequence at the same time.The performance of the decoding algorithm under different path values and different numbers of users is simulated experimentally.The results show that the performance of the JSCL decoding algorithm is improved with the increase of the path value,and the performance is the best at T≤4.Aiming at the uncoordinated massive random access problem in Gaussian multiple access channel,this thesis studies the low-complexity T-fold IRSA random access technology,and designs a scheme of T-fold IRSA based on Polar codes for the unsourced random access Gaussian channel.Based on T-fold IRSA,the length of transmission frame（?） is divided into two parts.The first part N_pis used to receive the preamble sequence,which contains the user’s transmission pattern information.The second part N_cis further divided into multiple slots.Active users randomly select different slots to repeatedly send data packet copies,and the repetition count obeys the given probability distribution L（x）.This thesis uses Polar code as the slot codes,and any the collision of the order up to T can be resolved by combining with the JSCL decoding algorithm.The interference caused by the successfully decoded user in other slots can be solved by combining the successive interference cancellation decoding algorithm between slots.The operation is carried out iteratively until the termination mechanism is triggered,and finally as much user information as possible can be recovered.The main goal of this thesis is to minimize the energy per bit E_b/N₀required to achieve the target packet loss ratioε.The experimental results show that,in the 4-fold IRSA scheme,using Polar code to replace LDPC code,its performance is improved about 1 to 2d B,and the gap between the finite block length boundary proposed by Polyanskiy is further narrowed.