射频非理想特性的基带补偿技术研究

发布时间：2018-03-03 15:36

本文选题：射频非理想特性　切入点：I\Q不平衡　出处：《电子科技大学》2017年博士论文　论文类型：学位论文

【摘要】：随着无线通信应用高度商业化,很多应用场景需要设计低成本低功耗的无线通信系统。然而,低成本系统面临各种挑战:限于成本,复杂的电路结构和昂贵的器件不再适用;同时,大规模生产的终端设备也不能对单台设备进行针对性的电路优化。这些因素导致低成本收发机存在严重射频非理想特性,造成系统性能严重下降。为缓解射频非理想特性造成的性能损失,发射端或接收端通常需要采用数字基带补偿或矫正算法。本文针对一些常见射频非理想特性展开研究,分析了射频非理想特性对系统性能的影响,并针对不同系统特点设计了矫正或补偿算法。本文具体的研究工作和创新点如下:首先,针对单载波系统接收机的同相与正交分量不平衡(In-phase\Quadrature Imbalance,I\Q不平衡),本文提出了基于格雷序列的I\Q不平衡时域补偿算法。接收机采用格雷序列对频选I\Q不平衡进行估计,利用序列自相关特性,极大地简化了估计过程。由于格雷序列在一些标准中被用于前导码或独特字(例如60GHz通信标准IEEE 802.15.3c和IEEE 802.11ad),该算法应用于采用上述标准的系统不需要改变帧结构,有效降低系统开销。仿真结果表明,该补偿算法在视距信道与非视距信道中都能消除I\Q不平衡影响。第二,针对发射机频选I\Q不平衡、载波泄漏以及带内失真,本文提出了自适应I\Q不平衡矫正和跟踪算法。每个支路逆响应直接通过自适应算法估计,避免了先估计I\Q支路响应再求逆滤波器的额外运算。同时,本文还提出一种迭代算法用于跟踪温度改变造成的I\Q不平衡参数变化。经过复杂度分析,该算法矫正效率较高。仿真结果对上述算法针对各类射频非理想特性矫正的有效性进行了验证。第三,本文定量分析了正交频分复用(Orthogonal Frequency Division Multiplexing,OFDM)差分空时编码(Differential Space-time Coding,DSTC)系统中,发射机和接收机I\Q不平衡的性能影响,并提出了一种在接收端联合补偿发射机和接收机I\Q不平衡的算法。利用判决导向的广义线性自适应算法,有效地缓解了I\Q不平衡对系统性能的影响。另外,本文还提出了一种基于I\Q不平衡参数估计的通用补偿算法。与广义线性补偿算法相比,该算法加快了收敛速度,提升了在快变信道中的补偿性能,并能分离收发端的I\Q不平衡参数。上述两种算法都利用差分编码规则进行自适应估计与补偿,不需要额外的训练序列或已知导频信号,节约了系统的开销,补偿性能经过了仿真验证。最后,本文定量分析了DSTC-OFDM系统中维纳相位噪声对系统性能的影响。在不使用额外训练序列或者已知导频信号的条件下,本文提出了一种判决导向算法估计并抑制相位噪声的公共相位误差。为增强对较高水平相位噪声的抑制能力,本文还提出了一种基于分组的判决导向估计算法。该算法利用分组规则将一个OFDM符号内的子载波分为两组,每组分别进行判决导向估计,再将估计结果进行组合和选择,有效地规避了判决错误,保证估计准确性。本文对上述两种算法进行了仿真验证,结果表明判决导向估计能够补偿较低水平的相位噪声;在严重相噪下,分组判决导向估计补偿性能较好。
[Abstract]:With the development of wireless communication applications highly commercialized, many applications that require wireless communication system design of low cost and low power consumption. However, the low cost system faces many challenges: limited to cost, complex circuit structure and expensive device is no longer applicable; at the same time, the terminal equipment of mass production can not be a single device for circuit optimization of. These factors lead to low cost RF transceiver has serious non ideal characteristics, resulting in a serious decline in system performance. In order to alleviate the non ideal characteristics of RF performance loss caused by the transmitter or receiver usually need to use digital baseband compensation or correction algorithm. In this paper, aiming at some common non ideal characteristics of RF, analyzes the influence of non ideal frequency the characteristics of the performance of the system, and the design of the correction or compensation algorithms for different system characteristics. The specific research work and innovations such as : first, according to the in-phase and quadrature components of single carrier receiver imbalance (In-phaseQuadrature Imbalance, IQ imbalance), based on Gray series IQ imbalance compensation algorithm. The receiver adopts time domain sequence of Gray frequency selective IQ imbalance estimation, using sequence autocorrelation, which greatly simplifies the estimation process as the Gray sequence in some standard is used in the preamble or unique words (such as 60GHz IEEE 802.15.3c and IEEE 802.11ad communication standard), the algorithm is applied to the system by using the standard does not need to change the frame structure, effectively reduce the system overhead. The simulation results show that this compensation algorithm can eliminate the effect of IQ imbalance in the horizon channel with the NLOS channel. In second, the transmitter frequency selective IQ imbalance, and carrier leakage in band distortion, this paper presents an adaptive IQ unbalance correction and tracking algorithm. Each branch is estimated by direct inverse response adaptive algorithm, avoid the first estimate of IQ branch in response to additional operations inverse filter. At the same time, this paper also puts forward an iterative algorithm for tracking the temperature change caused by the IQ imbalance parameters change. Through the complexity analysis, the algorithm has higher efficiency. The simulation results of the correction of the algorithm for all kinds of the effectiveness of non ideal characteristics of RF correction is verified. Third, this paper quantitatively analyzes the orthogonal frequency division multiplexing (Orthogonal Frequency Division Multiplexing, OFDM) differential space-time encoding (Differential Space-time, Coding, DSTC) system, affects the performance of the transmitter and the receiver IQ imbalance, and proposes a terminal joint compensation transmitter and the receiver IQ imbalance in the receiving algorithm. The generalized linear adaptive algorithm using decision oriented, effectively alleviate the imbalance of system performance IQ The effect. In addition, this paper also proposes a general IQ imbalance compensation algorithm based on parameter estimation. Compared with generalized linear compensation algorithm, this algorithm has faster convergence speed, improve the compensation performance in fast time-varying channel, and can separate transmitter and receiver IQ imbalance parameters. The two kinds of algorithms using the differential encoding rules for adaptive estimation and compensation, do not need additional training sequence or known pilot signal, saving the cost of the system, the compensation performance after simulation. Finally, this paper quantitatively analyzes the effect of Wiener phase noise on performance of DSTC-OFDM systems. Without the use of additional training sequences or known guide frequency signals, this paper proposes a decision directed estimation algorithm and phase noise suppression of common phase error. In order to strengthen the capability of restraining the high level of phase noise, this paper also proposes a Group decision directed estimation algorithm based on the algorithm. According to the grouping rules will be an OFDM symbol subcarriers are divided into two groups, each group of decision directed estimation, the estimation results of the selection and combination, effectively avoid decision errors, ensure the estimation accuracy. This paper carried out the simulation for the two the results show that the decision directed estimation algorithm, the phase noise can be compensated at lower levels; in severe phase noise, group decision directed estimation and compensation performance is better.

【学位授予单位】：电子科技大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TN929.53

【相似文献】