Description
The presence of both multiple-access interference (MAI) and intersymbol interference(ISI) constitutes a major impediment to reliable communications in multipath code-divisionmultiple-access (CDMA) channels. Many multiuser detectors in the past have been developedbased on novel linear and nonlinear interference suppression technique , which make useof both soft interference cancellation and instantaneous linear minimum mean-square errorfiltering. All of them are required knowledge of the spreading sequences , the initial delaysand multipath channels of the desired users. therefore, the knowledge of channel parametersare essential to achieve an efficient detection of the transmitted signal. The main purpose of this dissertation is to propose an advance novel solution to theproblem of joint multiuser detection, channel estimation and data detection for the uplinkof a multiuser DS-CDMA system exploiting. The devised algorithm from the application ofVariational Bayes Methods and Minimum Kullback-Leibler (MKL) techniques. MinimumKullback-Leibler (MKL) methods allow to efficiently numerical conditional distributionalprobability and parameter approximation of all the unknown parameters (i.e., data symbols,channel response of each users) from their conditional posterior distributions and then tocompute their estimates by condition mean estimator. In addition, they are insensitive,unlike the Expectation maximization (EM) technique, to the choice of initial conditions andconsequently perform robustly in quickly changing communication scenarios. The proposedMKL Bayesian multiuser dectector, being soft-input soft-output in its nature, can be exploitedfor iterative processing in a coded system. The simulation results have demonstrated relevant merit of the optimum MinimumKullback-Leibler (MKL) multiuser detector over multipath code-division multiple-access(CDMA) channels in the system with the unknown the initial delays and multipath channelsof the desired users. A few pilot symbols are used by any users for channel parametersat initial. The simulation results confirm performance properties of proposed MinimumKullback-Leibler (MKL) multiuser detector and show the advantages of the novel design over others.
Citations (0)
No citations found
Mentions (0)
No mentions found
Metrics Over Time
Publication Details
Subfield
Artificial Intelligence
Field
Computer Science
Domain
Physical Sciences
Confidence Score
51%
Source
Scholar Data Model