论文部分内容阅读
针对长码直接序列扩频码分多址(DS-CDMA)信号的盲解扩,在信号模型分析的基础上,提出了一种基于可逆跳跃的马尔可夫链蒙特卡罗(RJ-MCMC)扩频码和信息序列联合估计算法。该算法分别对建立的联合后验分布模型进行迭代抽样,并有效地在不同维数的子空间中跳转,从而构造一条马尔可夫链,使其平稳分布为待估参数的后验分布,最终拼接得到每个用户的完整扩频序列和信息序列估计。仿真结果表明:该方法在迭代二十几步时达到收敛;并且在功率相同和不同条件下,当信噪比(SNR)大于-9dB时,估计序列与真实序列的相似度均超过0.95,信息序列的误码率低于0.01;同时算法对不同用户个数和不同调制样式具有较强的适应性,与Fast-ICA算法相比,估计性能平均提高了约3dB。
Aimed at the blind despreading of long code direct sequence spread spectrum code division multiple access (DS-CDMA) signals, a Markov chain Monte Carlo (RJ-MCMC) based on reversible jump was proposed based on the signal model analysis. Joint Estimation Algorithm of Spread Spectrum Code and Information Sequence. The algorithm iteratively samples the established joint posteriori distribution model and jumps efficiently in subspaces of different dimensions, so as to construct a Markov chain, making it a posterior distribution of the estimated parameters. The final splicing results in a complete spread spectrum sequence and information sequence estimate for each user. The simulation results show that the proposed method converges more than 20 steps in iteration, and the similarity between the estimated sequence and the real sequence exceeds 0.95 when the signal-to-noise ratio (SNR) is greater than -9dB under the same power and different conditions. The information The bit error rate of the sequence is less than 0.01. Meanwhile, the algorithm has strong adaptability to different number of users and different modulation styles. Compared with the Fast-ICA algorithm, the estimated performance is improved by about 3 dB on average.