Research On Channel Estimation Algorithms For Low Complexity Communication Systems

Regardless of the type of channel model used,the common challenge of different communication systems at the receiver is to provide as perfect channel state information（CSI）as possible for accurate detection,signal demodulation,decoding,and other baseband module applications.Therefore,the channel estimation module has always been a research hotspot in communication systems.As a semi-blind channel estimation algorithm,the expectation maximization（EM）channel estimation algorithm can achieve better performance within an acceptable complexity range,is a good compromise between performance and complexity.There is a lot of optimization work for the EM algorithm,but there is a lack of simulation applications built on actual scenarios,and the existing simplified solutions have limited reduction in complexity.Based on this,this paper first discusses the relationship between the performance and complexity of the algorithm of the individual channel estimation module.Through the numerical analysis of the simulation performance of the EM channel estimation algorithm applied in the indoor office scene and a new perspective on the derivation of the EM algorithm formula,the EM-MAX algorithm is proposed.Simulation experiments show that the proposed EM-MAX algorithm can reduce the complexity of the original EM algorithm by 88.6%while the performance loss is within 1 d B.Furthermore,the research hotspot on channel estimation algorithms has turned to the research on reconfigurable smart surfaces（RIS）assisted large-scale millimeter-wave systems.At present,there have been a number of research works on the establishment of joint channel models for RIS-assisted wireless communication processes.On the basis of these joint channel models,channel estimation al-gorithms based on compressed sensing（CS）can be applied.However,most of the channel estimation algorithm research for RIS-assisted sparse channel model is to create a pilot transmission protocol to reduce the system pilot overhead,and there is less work for joint channel estimation and detection or decoding.Further combined with the role of channel estimation in the multi-module joint system,in this paper,an improved and simplified scheme is proposed for the joint detection module of chan-nel estimation,and a dual-module early-stop decision cooperative algorithm is proposed for the joint channel estimation and decoding system.On the one hand,for the research of the channel estimation module joint detection module system,we use the first round of channel detection results as the reference signal at the data,and input the original pilot together into the OMP channel estimation algorithm,so as to obtain an improved joint channel estimation algorithm.Channel estimation and detection algorithms.Simulation experiments show that the proposed improved joint channel estimation and detection algorithm has a performance improvement of 5 d B compared with the original algorithm when the MSE performance reaches the order of 10^-4.In addition,we further use the multi-parameter early stop decision algorithm to simplify the two OMP channel estimation modules of the improved system.The simulation results show that the proposed early stop scheme can greatly reduce the system complexity.On the other hand,for the performance and complexity trade-off between the channel estimation module and the decoding module,we use the iterative channel estimation and decoding algorithm to explore the effect of single module complexity improvement on the system performance.By observing the FER increment caused by the increase of the number of iterations of a single module,we find the inherent law of the interaction between the channel estimation and decoding modules,and then propose an early stopping decision scheme suitable for the channel estimation and decoding modules,respectively.Further,we propose an early-stop decision coordination algorithm under the nesting of double-iterative loops,which realizes the automatic decision of the number of iterations of the two modules and improves the system efficiency to the greatest extent.The simulation results show that the joint channel estimation and decoding system based on the early-stop decision coordination algorithm can adaptively obtain the optimal combination of the number of iterations of the two modules under different signal-to-noise ratio（SNR）conditions,which greatly reduces the complexity of the system and achieves the most efficient use of system resources and realizes the optimal allocation of system resources.