-
Codes for Optical CDMA
Tue, Nov 14, 2006 @ 03:30 PM - 04:30 PM
Ming Hsieh Department of Electrical and Computer Engineering
Conferences, Lectures, & Seminars
SPEAKER: Mr. Reza Omrani, Communication Sciences Institute, University of Southern CaliforniaABSTRACT: There has been a recent upsurge of interest in applying code division multiple access (CDMA) techniques to optical networks. This interest is in part due to the increase in security afforded by OCDMA as measured for instance, by the increased effort needed to intercept an OCDMA signal, and in part due to the flexibility and simplicity of network control afforded by optical-CDMA (OCDMA). There are two main approaches to code design for OCDMA systems. The first approach uses direct sequence encoding, which employs {0,1} sequences with good correlation properties as code sequences, and the data sequence modulates the code sequence simply by switching it on or off. These may be termed as 1-D optical orthogonal codes (OOC) since the code sequences are only associated with the time dimension. It has recently been recognized that in order to bring down the required chip rate to within practical limits, it is desirable that 2-D {0,1} codes be used in which the code sequence consist of a 2-D pattern in which the second dimension corresponds to wavelength.The second OCDMA approach is via phase encoding in which the code sequences are collections of complex numbers of unit magnitude with each entry associated to a carrier of different wavelength. The phase of an element in the λ-th sequence corresponds to the phase of the λ-th carrier. The focus of our work is on efficient code design for OCDMA systems under both direct sequence and phase encoding approaches.We first introduce some new bounds on the size of 1-D and 2-D OOCs. Subsequently, the focus is on explicit constructions for 2-D wavelength-time OOCs. We introduce four major constructions for wavelength-time OOCs which include a method to map 1-D OOCs to 2-D OOCs, a method based on Reed-Solomon Codes, a method which concatenates a constant weight code with a Reed-Solomon based 2-D OOC and finally a function-plot method in which the values of an appropriately-chosen function are used to derive the 2-D codes. The functions used in function plot construction include polynomial functions and rational functions.A major drawback of OCDMA systems is their low spectral efficiency. In this work we explore a modulation scheme for OCDMA systems which has the potential for increasing the spectral efficiency manyfold. We term this scheme code-cycle modulation (CCM). Under this modulation scheme, different cyclic shifts of the code sequence assigned to each user are used to transmit M-ary information. While this means of modulation was known earlier in the literature, most prior modulation schemes needed M different receiver units to recover the data resulting in increased complexity and power. In this work we provide a novel receiver architecture with significantly reduced power requirements and complexity.Following this, we turn our attention to phase-encoded OCDMA. We first derive a mathematical model for the output of this system and based on this model we introduce a metric to design code sequences for asynchronous transmission. Then, a connection between the phase-encoding sequence design problem and the PMEPR (peak to mean envelope power ratio) problem which arises in OFDM transmission is established. We construct a family of phase sequences which are based on the theory of generalized bent functions and with properties desirable for asynchronous phase encoding OCDMA systems.BIO: Reza Omrani received his B.S. degree from the University of Tehran, Tehran, Iran, and his M.S. degree from the University of Southern California (USC), Los Angeles, both in electrical engineering, in 1999 and 2002, respectively. He did his PhD dissertation under the supervision of Professor P. Vijay Kumar at the Communication Sciences Institute, Department of Electrical Engineering-Systems, USC. His research interests include low-density parity-check (LDPC) codes, network coding, combinatorics and signal design for good correlation properties.Host: Michael J. Neely, mjneely@usc.edu
Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248
Audiences: Everyone Is Invited
Contact: Mayumi Thrasher