Research On High-precision Synchronization Algorithm Based On 5G-TSN Systems

With the continuous development of Industry 4.0,there are new requirements for highprecision time synchronization,and it is necessary to find some emerging technologies to complete rapid synchronization.The combination of Time Sensitive Networking(TSN)and fifth-generation mobile communication(5G)can bring significant benefits to industrial scenarios in various vertical fields.5G-TSN can provide both ultra-reliable and low-latency communication,as well as High-precision synchronization and timing mechanism.Time synchronization in the 5G time domain in 5G-TSN is an important part.In the process of time synchronization,the base station needs to distribute time information to the user equipment(UE).The UE needs to perform downlink synchronization and estimate the propagation delay to ensure the realization of high-precision synchronization.Therefore,this paper focuses on the 5G downlink synchronization algorithm in the low SNR and large frequency offset environment,as well as the propagation delay estimation algorithm suitable for the line-of-sight/non-line-of-sight(LOS/NLOS)environments.The most difficult stage in downlink synchronization is the primary synchronization signal(PSS)detection process.Since the UE does not have any prior information about the synchronization signal block(SSB)when detecting the PSS,it needs to scan according to the synchronization grid,which is more complicated.In order to achieve fast and highperformance downlink synchronization,this paper proposes a SSB detection algorithm based on convolutional neural network(CNN)and a multi-beam SSB joint timing synchronization algorithm.The SSB detection algorithm based on CNN can realize SSB detection in wireless environments such as low signal-to-noise ratio or large frequency offset,and locate the signal segment occupied by SSB.Then,the SSBs carried by multiple beams are combined in the segment to perform PSS timing information detection and frequency offset estimation,making full use of the correlation between beams to achieve efficient detection,and avoid invalid search caused by direct point-by-point sliding correlation.The simulation results show that the proposed algorithm based on CNN has good SSB detection accuracy,can effectively resist the influence of low signal-to-noise ratio and large frequency offset in the synchronization process,and maintain a high PSS timing synchronization success rate.In the 5G time domain in 5G-TSN,the UE needs to perform delay compensation after receiving the system message block(SIB)timestamp.Accurate estimation of the propagation delay is a key factor to achieve device-level synchronization.Aiming at the propagation delay estimation in 5G-TSN access network synchronization,this paper proposes a propagation delay estimation algorithm to offset the non-line-of-sight path.Firstly,the proposed algorithm uses the imported vector machine(IVM)based on the feature selection strategy to identify the LOS/NLOS environment,and counteracts the non-line-of-sight path if the current environment is the NLOS environment.Furthermore,perform eigenvalue decomposition on the covariance matrix of the signal,and the density clustering(DBSCAN)algorithm with noise is used to cluster the signal subspace and noise subspace unsupervised.The multipath number can be obtained according to the number of samples of the cluster in the signal subspace.Finally,the propagation delay estimation is realized by using the improved multiple signal classification(MUSIC)algorithm.Simulation results show that the proposed algorithm can more accurately identify the NLOS environment,estimate the number of multipaths and finally estimate the propagation delay,and its performance is better than the existing algorithms.In addition,the effects of bandwidth and cancellation times on the performance of the algorithm are also explored in the simulation.