Key Technologies On Low-cost Advanced Receivers For 6G

To meet the continuously increasing demands of mobile communications,6th generation(6G)mobile communication systems are expected to support ultra-high data rate up to several gigabits per second.Increasing the size of antenna array and expanding bandwidth are still the key technologies to realize the vision of 6G.However,receivers implemented with the conventional fully digital and high-precision hardware architecture gives rise to vital practical challenges including massive baseband data to be processed,and prohibitively high hardware cost,power consumption and computational complexity.Regarding this,the design of low-cost highefficiency receiver architectures becomes one of the key issues of 6G.Low-resolution quantized receiver,hybrid receiver and decentralized receiver are promising low-cost receiver architectures.This dissertation focuses on the development of key techniques including data detection,channel estimation and configurations of hybrid combiners for these low-cost architectures.Firstly,an iterative data detector is developed based on Generalized Turbo(GTurbo)framework for Orthogonal Frequency Division Multiplexing(OFDM)system with low-resolution Analog-to-Digital Converters(ADCs).The state evolution expression of the proposed GTurbo detector is derived,from which the analytical symbol error rate expression is obtained.On the basis of the derived state evolution expression,a power allocation scheme is proposed aiming to minimize the average symbol error rate.The state evolution expression matches the saddle point of average free energy,demonstrating that the proposed GTurbo detector asymptotically achieves the fundamental limit of Minimum Mean Square Error(MMSE)estimator.Simulation results confirm the accuracy of the state evolution analysis and illustrate that the performance of the proposed detector in conjunction with the proposed power allocation scheme is close to the optimal performance of the OFDM system with infinite-resolution ADCs.Then,the overall framework of quantized OFDM receiver is designed and the prototyping system is implemented.A channel estimation algorithm dedicated for the quantized OFDM systems is proposed by extending the GTurbo framework applied for optimal detector design.Specifically,a type of robust linear OFDM channel estimator is integrated into the original GTurbo framework,and its corresponding extrinsic information is derived to guarantee its convergence.Feasible schemes for automatic gain control,noise power estimation,and synchronization are also proposed.Combined with the proposed inference algorithms,an efficient quantized OFDM receiver architecture is developed.Furthermore,a proof-of-concept prototyping system is constructed and over-the-air experiments are conducted to examine reliability when transmitting quadrature phase shift keying(QPSK)with 1-bit quantization and 16-quadrature amplitude modulation(16QAM)with 2-bit quantization.The results of the numerical simulation and OTA experiment demonstrate that reliable OFDM transmission can be achieved when the signal-to-noise ratio is higher than approximately 8 d B for both cases.Subsequently,the Dynamic Metasurface Array(DMA)-based MIMO-OFDM receiver operating with bit-constrained ADCs is proposed.The mathematical model is presented which accounts for the configurable frequency selective profile of its metamaterial elements of the DMA,resulting in a frequency-dependent configurable hybrid receiver structure.Then the task-based quantization framework is exploited to approximate the MSE expression of OFDM signal recovery in the presence of bit-constrained ADCs.Based on the matrix quadratic transformation,the MSE minimization problem is transformed into a constrained quadratic optimization problem.Correspondingly,alternating algorithms are proposed for different types of DMA weights based on matrix fractional programming.Numerical results show that the DMA-based receiver is capable of accurately recovering OFDM signals,demonstrating the potential of DMAs in realizing high performance massive antenna arrays of reduced cost and power consumption.Finally,Expectation Propagation(EP)-based decentralized detector for extra-large-scale massive Multiple Input Multiple Output(MIMO)systems is proposed,in which baseband data from disjoint subsets of antennas,called subarray,are distributed into parallel processing procedures coordinated by a central processing unit.The vector-valued factor graph is applied to represent the a posteriori distribution of the considered scenario and messages propagated among nodes on the factor graph are updated based on EP principle.Analytical results confirm the convergence of the proposed detector and that the proposed detector asymptotically approximates the fundamental limit of MMSE performance under certain conditions.In addition,additional strategies are proposed for further reducing the complexity and overhead of the information exchange between parallel subarrays and the central processing unit to facilitate the practical implementation of the proposed detector.Simulation results demonstrate that the proposed detector achieves numerical stability within five iterations and outperforms its counterparts.