多进制部分响应Multi-h CPM的软输出维特比算法

为降低多调制指数连续相位调制(Multi-h CPM)信号解调和级联译码的复杂度和时延,提出了一种适用于多进制部分响应Multi-h CPM信号的低复杂度软输出维特比(LC-SOVA)算法。针对倾斜相位(TP)+频率脉冲截断(FPT)低复杂度Multi-h CPM解调,在维特比检测中对每个状态存储并不断更新符号软信息,通过舍弃度量较小的路径来降低更新软信息的运算量,且软信息更新过程可以并行处理,复杂度降低一半,时延与维特比算法等价。仿真结果表明,所提LC-SOVA算法具有通用性,适用于多进制部分响应Multi-h CPM信号的软信息计算,在级联低密度奇偶校验码(LDPC)系统中,该算法与Max-log-MAP算法相比信噪比损失几乎可以忽略不计。     

软扩频Multi-h CPM信号设计方法及其性能

为解决传统多调制指数连续相位调制(Multi-h CPM)实现复杂度高和调制指数同步困难的问题,提出一种软扩频Multi-h CPM信号设计方法。利用单调制指数软扩频获得Multi-h CPM类似的相位轨迹和性能增益,伪码捕获的同时实现调制指数同步;推导了软扩频Multi-h CPM欧氏距离表达式,搜索获得了h=5/7时的最小欧氏距离,与直接序列扩频单调制指数CPM(DSSS Single-h CPM)信号相比,至少可获得0.5 dB的欧氏距离增益。基于极大似然序列检测(MLSD)的同步解扩解调算法,仿真了h=5/7的软扩频Multi-h CPM误码性能,与DSSS Single-h CPM信号相比,在BER=10–5时,可获得1.5 dB性能增益,且性能优于美军WGS系统和ARTM Tier II系统采用的Multi-h CPM信号。     

基于状态空间分类的Multi-h CPM序列检测算法

考虑到Multi-h CPM信号序列检测复杂度较高,难以工程实现,在最大似然序列检测的基础上,针对该算法后端复杂的网格状态,提出基于状态空间分类的Multi-h CPM序列检测算法。通过引入倾斜相位,避免了区分调制指数分子为奇数和偶数的问题,利用状态空间分类合并的方法,大大简化了执行维特比算法所需的网格状态,并引入判决反馈原理,提高了误码性能。最后对ARTM TierⅡ信号进行仿真,结果表明,该算法可将网格状态减少为最大似然序列检测算法网格状态数的1/32,在10-5误码率下的性能损失仅为0.6 dB。     

一种新颖的Multi-h CPM信号符号速率盲估计算法

针对非合作条件下Multi-h连续相位调制(CPM)信号符号速率的盲估计问题,提出了一种新颖的预滤波二次谱算法,通过对接收信号滤波然后搜索其二次谱的峰值实现Multi-h CPM信号符号速率的估计,并利用CPM信号的PAM展开表达式解释了算法的原理.仿真实验结果表明,该算法是无偏的,与基于循环谱分析的算法相比具有运算量小的优点,并且适用于各种类型的Multi-h CPM信号,同时不需要载波频率和调制指数的先验信息,具有很强的鲁棒性.     

大动态PCM/FM遥测信号的非相干解调

针对大动态低码率PCM/FM遥测信号的可靠接收问题展开研究,给出了一种基于连续相位调制（CPM）信号原理的载波频偏矫正及差分序列检测的非相干接收方案。针对限幅鉴频、多符号检测及差分序列检测等3种非相干解调算法,介绍了各自应用于PCM/FM信号的算法原理,并比较了3种算法的解调性能及对大动态环境的适应能力。计算机仿真表明,在无残留频偏情况下,2符号相位差分序列检测性能最好;但在0.05倍码速率的残留频偏下,1符号相位差分序列检测的性能更好,优于2符号相位差分序列检测、限幅鉴频检测和多符号非相干检测。对PCM/FM信号的讨论包含了几种典型的非相干解调,对大动态环境下的PCM/FM接收有重要的参考价值。     

串行级联多指数连续相位调制的迭代检测

针对多调制指数连续相位调制(Multi-h CPM)信号给出了卷积编码、随机符号交织器串行级联系统的迭代检测,研究了针对Multi-h CPM信号的逐符号最大后验概率(MAP)检测算法及不同系统参数下的迭代检测性能。最后提出在频率脉冲截断(FPT)的基础上使用奇异值分解(SVD)的联合降复杂度算法,有效简化了部分响应多指数CPM信号的状态转移网格以及接收端的匹配滤波器。在高斯白噪声信道下,针对第二代遥测新体制(Tier2)信号进行仿真,与相同码率及码长的低密度奇偶校验(LDPC)编码系统相比,串行级联系统有着3.5 dB的性能增益。最后采用本文提出的联合降复杂度算法,能够将Tier2信号的网格状态数由256简化到64,匹配滤波器个数由64降到2,而由此带来的迭代检测性能损失却可以忽略不计。     

连续相位调制的非相干复合网格解调算法

该文提出了一种基于复合网格处理的CPM信号非相干解调算法。该算法将CPM信号的基本网格图和相位旋转网格图合并组成复合网格图,并在复合网格图的单元网格之间引入状态耦合转移,实现了对CPM信号初始相位和频偏的有效跟踪。仿真表明,当单元网格数达到4个以上时,该文提出的非相干解调算法性能非常接近相干解调,且具有较好的频偏跟踪性能,对于载波恢复困难的盲解调情况具有实用意义。     

微波通信中连续相位调制的非相干检测算法

提出了一种适合微波通信传输使用的多进制连续相位信号的频偏校正2码元差分检测算法.该算法利用2码元差分信息提取分支度量,进行维特比检测.同时利用判决反馈对微波通信中的载波频偏进行校正.通过对接收信号的复指数分解,分支度量的计算量得到了很大的简化.仿真结果表明,该算法的性能优于1码元差分检测.当接收信号存在载波频偏时,该算法仍然具有较好的检测性能.     

基于FPGA的Multi-h CPM高速解调系统设计与实现

多指数连续相位调制(Multi-h Continuous Phase Modulation,Multi-h CPM)作为一种高效的调制方式,具有良好的频谱效率和抗干扰能力,在卫星通信领域有着广泛的应用价值。但是Multi-h CPM的解调算法计算复杂度较高,为了实现高速数据传输能力,需要较高的硬件资源开销。因此基于现场可编程门阵列(Field Programmable Gate Array,FPGA)平台的高并行度数据处理能力,设计了一种Multi-h CPM高速解调系统,采用了软输出维特比算法(Soft Output Viterbi Algorithm,SOVA)实现了Multi-h CPM信号解调,并通过早迟门和升降频门实现了时频同步功能。为了降低算法的计算复杂度,本系统引入了倾斜相位(Tilted Phase,TP)算法和频率脉冲响应截断(Frequency Pulse Truncation,FPT)算法减少了Viterbi过程的网格状态数,降低了系统的资源开销。通过仿真和实际测试,验证了所提出的解调系统的有效性,并表明该系统具有良好的解调性能和系统稳定性。     

基于PSP均衡技术的CPM检测算法

针对时变频率选择性衰落信道,研究了连续相位调制（CPM）信号的逐幸存检测算法。该算法在未知信道状态的条件下,利用训练序列对信道参数进行初始估计。在对CPM信号进行Viterbi解调过程中,采用PSP技术实现信道的无延时跟踪。基于频域均衡的CPM检测算法虽然可以有效抗多径干扰且计算复杂度较低,但不能对时变信道进行跟踪。仿真结果表明,在时变多径信道下,基于PSP均衡的CPM检测算法能有效地进行信道参数估计,比频域均衡算法具有更好的误码性能。     

Reduced phase-state detection based on virtual modulation indexes for Multi-h CPM

.Simplified detection with reduced phase-state based on virtual modulation indexes was designed for Multi-h CPM.Multiple sets of modulation indexes different from the transmitter were constructed in the receiver,and the per-survivor processing was used to solve the mismatch caused by the inconsistency of the modulation indexes between the receiver and the transmitter.When the denominator of virtual modulation index was much smaller than the original denominator of actual transmitted signal,the phase states of coherent sequence detection could be effectively simplified,thus the complexity of coherent detection for Multi-h CPM could be reduced.The trellis could be simplified further via combining with the frequency pulse truncation(FPT) used for simplifying correlation states.The simulation results show that for advanced range telemetry (ARTM) CPM signal,various simplified receiving schemes based on virtual modulation indexes can be designed.When the virtual modulation indexes are set to $\left[\frac{\text{3}}{\text{4}}\text{,}\frac{\text{2}}{\text{4}}\right]$,the number of trellis states can be reduced from 512 (in optimal detection) to 64,but the performance loss is less than 0.05 dB.Further combined with FPT algorithm,the number of trellis states can be reduced to 16,while the performance loss is only 0.75 dB.     

Multiple differential detection of continuous phase modulationsignals

A sequence estimation algorithm for the differential detection of the continuous phase modulation (CPM) signals, yielding significant gains in BER performance and with considerable resistivity to fading, is introduced. These advantages, along with the reduced hardware complexity, low cost, and fast synchronization which characterize the differential detector, make the proposed receiver useful for land mobile radio and mobile-satellite communications. The new receiver is based on multiple differential detection. The multiple differential detection strategy provides the decoder with more information regarding the transmitted data and applies a noise decorrelation process on the received signal, useful to the sequence estimation. The algorithm is derived in a general form, and can be applied on any CPM scheme, with any degree of complexity. The authors have evaluated the receiver for two of the most popular CPM schemes, the tamed frequency modulation (TFM) and Gaussian minimum-shift keying (GMSK) (with B₁ T=0.25), in the presence of additive white Gaussian noise (AWGN) and Rician fading. The BER performance evaluation results indicated significant gains and considerable reduction of error floors. In AWGN improvements close to 9 dB have been verified     

LTE系统中一种串行的整数倍频偏和扇区检测算法

在LTE系统中,用户终端( UE )在开机后首先会进行小区搜索。针对传统的小区搜索方案中整数倍载波频偏( ICFO)和扇区ID是通过利用主同步信号( PSS)进行联合检测这一问题,提出了一种串行的整数倍频偏和扇区ID估计算法。该算法利用PSS的对称性,将ICFO和扇区ID进行分开检测,通过对接收的频域PSS进行归一化差分相关,消除了信道的影响,从而增强了检测性能。将联合估计算法和提出的串行估计算法进行仿真对比,结果表明提出的算法相较于传统方法可以取得更好的检测精度,并且运算复杂度仅为传统方法的1/3。     

一种CPM信号参数盲估计方法

针对非协作通信下CPM信号参数盲估计问题,提出了一种系统的有序的盲估计方法,实现了对CPM信号的载频、符号速率、调制指数和调制阶数等参数的精确估计。利用CPM信号的循环平稳特性,采用了基于周期谱的符号速率估计方法,构建了代价函数,采用联合估计法对调制指数和频偏实现了三维的精确搜索,最后实现了基于循环平稳性的CPM信号调制阶数的准确识别。理论分析与试验仿真表明,新方法具有较高的正确估计概率,且在低信噪比下也能达到较好的效果。     

Carrier Synchronization for Very Low SNR

The Turbo decoding performance will suffer serious degradation under low signal-to-noise ratios (SNR) conditions for the reason of residual frequency and phase offset in the carrier. In this paper, an improved residual carrier frequency offset estimation algorithm based on u priori probability aided (APPA) phase estimation is proposed. A carrier synchronization loop that combines the iterative turbo decoder and the phase estimator together is constructed, where the extrinsic information obtained from the Turbo decoder is used to aid an iterative phase estimation process. The simulation results show that the algorithm performs successfully under very low SNR conditions (for example, less than -7.4 dB) with large frequency offset and phase error and the performance of this algorithm is very close to the optimally synchronized system.     

一种基于APPA的码辅助载波同步算法

在低信噪比环境下,小的频偏和相偏的存在会使Turbo编码系统的译码性能恶化,所以必须结合迭代译码系统,对信号的残留频偏和相差进行估计。该文提出一种改进的残留频偏载波相位估计算法后验概率辅助(APPA)相位估计,该算法将译码器输出的外信息用于辅助迭代的相位估计。环路滤波器将相位误差信号转换成控制信号,控制数控振荡器的输出,这样可以得到待估计的相位误差。仿真表明,在极低信噪比下(比如,SNR-7.8 dB)该算法在同时存在频偏和相偏的时候能正常工作,其性能非常接近理想同步条件下的性能。