Optimum Codes for Overloaded CDMA Systems

Please enable javascript in your browser.

Heidari Khoozani, Mohsen | 2013

590 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 44480 (05)
University: Sharif University technology
Department: Electrical Engineering
Advisor(s): Marvasti, Farrokh
Abstract:
CDMA is one of the main multiple access techniques. The aim of this work is to propose optimum signature codes to improve the performance of such systems. First, we introduce criteria such as Sum Channel Capacity and Bit Error Rate (BER) to measure the performance of a CDMA system. Also a set of measures named as Distance criteria are proposed to reduce the computational complexity. Various optimization methods are used to optimize these criteria such as Genetic Algorithms and the Particle Swarm Optimization techniques. Then, several sub-optimum signature matrices are derived based on different factors and criteria to improve the performance of Overloaded CDMA systems in terms of the sum channel capacity and BER. Furthermore, we propose a new method to enlarge small-scale signature matrices to achieve sub-optimum large dimension matrices.
The aforementioned systems are binary input with dissimilar signature matrices. Two different types of matrices, i.e., binary and real valued matrices are considered and optimized
Keywords:
Sum Channel Capacity ; Code Devision Multiple Access (CDMA) ; Performance Criteria ; Signature Matrixes

No TOC