Font Size: a A A

Research On Communication Signal Joint Processing Techniques

Posted on:2014-02-25Degree:DoctorType:Dissertation
Country:ChinaCandidate:G WangFull Text:PDF
GTID:1108330482479108Subject:Military information science
Abstract/Summary:PDF Full Text Request
The rapid development of modern communications, under the premise of reliable transmission, needs like lower transmit power, higher data rates and higher spectral efficiency are constantly emerging. Therefore high-order modulation, efficient channel coding and other advanced signal processing technologies are required in modern communication systems. Complex wireless environment, continuously reduce the signal power and gradually decreasing Euclidean distance, make the receiving technology faces more severe challenges. The study of reliable reception technology for low SNR signals(Close to or lower than the existing receiver threshold) has been increasingly urgent. We need a new processing architecture and signal processing technologies. Therefore, the joint processing technology that using the correlation of different data streams or the information of the same data streams between different levels to get the processing gain has highlighted its greater advantage, and become an important development direction of modern digital receivers.Researches works of this paper are some joint processing techniques of communication signals, by analyze and utilize the posteriori information to improve processing performance. Based on the ML joint multi-parameter processing technology, this paper analyzes joint symbol detection and synchronization, joint symbol detection and channel estimation, and a probability constraints symbol information synchronization algorithm. The main work and innovations of this paper can be summarized as the following aspects:1、For the convergence issues of expectation-maximum algorithm, the thesis analysis the relationship of the rate of convergence of the algorithm and the parameter estimation Cramer-Rao Bound from the theoretical, and proposed the fast convergence parameter estimation algorithm based on EM. Theoretical proof and simulation results show that the algorithm does not reduce the estimated performance while effectively improve the convergence rate. For unknown transmitted symbols synchronization parameter estimation problem, starting form ML estimates of missing data,a synchronization parameter estimation algorithm is derived based on the EM algorithm. On the convergence rate of the EM algorithm, this paper analysis the convergence performance. The relationship between parameter estimation CRB with the convergence rate of the EM algorithm is proved, that is lower parameter estimation CRB, the faster convergence rate of the EM algorithm. Based on a fast convergence method of parameter estimation is proposed, which analyze and utilize the posteriori information, and reduce the uncertainty of the transmission symbol. This paper also proved the accelerating convergence mechanism of the improved algorithm. That is reduces the entropy of the missing data. And it also proves that the modified algorithm still converges to the likelihood function before correction. That means the correction is not affected to the performance of the parameter estimation algorithm based on EM.2、Given the joint estimation processing structure of symbol detection and carrier phase based on EM and alternating projection. Simulation and analysis the performance of the two processing structure. For the detection performance degradation problems was caused by unbiased estimate range smaller of the joint symbol detection and phase estimation, an adaptive phase expectation interval adjustment algorithm was proposed. Simulation results show that the algorithm can effectively enhance the performance of joint detection. For unknown synchronization parameters symbol detection problem, this paper gives a EM-based joint symbol detection and synchronization parameter estimation algorithm. In particular, for the symbol detection of a symbol with an unknown carrier phase, this paper proposes a joint symbol detection and phase estimation algorithm based on EM. The algorithm has changed the traditional optimum reception way of processing. For the traditional hierarchical approach, carrier phase estimation is required before symbol detection. The new algorithm using the EM method to detect symbols directly, without estimating the carrier phase. Since the algorithm does not require synchronization parameter estimation, it can greatly reduce the performance loss of the traditional hierarchical processing optimum reception method. After the convergence of the algorithm, symbol information is obtained by the decision, and the maximum likelihood(ML) estimation of carrier phase shift can be obtained according with the maximum a posteriori(MAP) criterion. So the paper proposes adaptive integration interval improvement algorithm. The algorithm uses the carrier phase estimation updated results to adjust the integral interval adaptive. Overcome the performance degradation caused by integration interval. Further, the traditional optimum reception algorithm, parameter estimation and symbol detection will be optimized separately, that may not get a global joint optimal solution. This paper studies this problem, and proposes a joint global multi-parameter processing algorithm based on alternating projection. Simulation results show that the algorithm outperforms the traditional optimum reception algorithm, which estimates the carrier phase using the non-data-aided method before symbol detection. The estimation performance can achieve the performance of pilot assisted symbol detection.3、For the symbol detection and channel estimation problem under the fading channels, introduce the particle filter algorithm, proposes a joint symbol detection and channel estimation algorithm based on EM-PF. The algorithm instead the continuous integration of channel with summation under limited value. Ensure the detection performance while improving the efficiency of the algorithm. Considering the optimum signal detection on optimum reception also depends on accurate channel parameter estimation, this paper extends the iterative method of EM algorithm to the frequency selective fading channel environment, The paper has designed a novel iterative receiver structure on joint symbol detection and channel estimation based o EM-PF, Since the channel parameter in the algorithm is a continuous variable, as the result, and it is difficult to obtain an analytic form of the integral expression. The paper using Monte Carlo methods by introducing the particle filter algorithm to obtain approximate integral value. By way of importance sampling, with a series of important samples values of channel parameters replaces the space approximately, the integral value is transformed into a limited space summation. 4、For different packet and different packet control fields of specific communication protocol exist hidden correlation and redundant information, a probability constraints symbol information synchronization algorithm is proposed based on the EM algorithm. Further the joint probabilistic constraint between multiple fields is modeled as unconscious encoding rules. We propose a factor graph parameter estimation algorithm based the probability constraint encoding method. Simulation results show that the algorithm can effectively improve the estimation under low SNR estimation. For different packet and different packet control fields of specific communication protocol exist hidden correlation and redundant information. The thesis modeled the hidden correlation and redundant information as a probabilistic constraint. Paper proposed the probability constraints symbol information synchronization algorithm can effectively utilize a priori redundant information of protocol structure fields to assist synchronization parameter estimation. Further the joint probabilistic constraint between multiple fields is modeled as unconscious encoding rules. We propose a factor graph parameter estimation algorithm based the probability constraint encoding method. Simulation results show that utilize the probability constraint information of transmission symbols, obtained performance improvements under low SNR estimation.
Keywords/Search Tags:Joint Process, Expectation Maximum, Parameter Estimate, Convergence Rate, Signal Detection, Channel Estimation, Particle Filter, Factor Graph
PDF Full Text Request
Related items