The New Jersey Institute of Technology's
Electronic Theses & Dissertations Project

Title:	Turbo space-time coded modulation : principle and performance analysis
Author:	Cui, Dongzhe
View Online:	njit-etd2001-061 (xii, 102 pages ~ 4.9 MB pdf)
Department:	Department of Electrical and Computer Engineering
Degree:	Doctor of Philosophy
Program:	Electrical Engineering
Document Type:	Dissertation
Advisory Committee:	Haimovich, Alexander (Committee chair) Tarokh, Vahid (Committee member) Soni, Robert A. (Committee member) Ge, Hongya (Committee member) Michalopoulou, Eliza Zoi-Heleni (Committee member)
Date:	2001-05
Keywords:	Turbo Code Space-Time--Coding Fading Channel Trellis Codes Maximum a Posteriori (MAP) Algorithm
Availability:	Unrestricted
Abstract:	A breakthrough in coding was achieved with the invention of turbo codes. Turbo codes approach Shannon capacity by displaying the properties of long random codes, yet allowing efficient decoding. Coding alone, however, cannot fully address the problem of multipath fading channel. Recent advances in information theory have revolutionized the traditional view of multipath channel as an impairment. New results show that high gains in capacity can be achieved through the use of multiple antennas at the transmitter and the receiver. To take advantage of these new results in information theory, it is necessary to devise methods that allow communication systems to operate close to the predicted capacity. One such method recently invented is space-time coding, which provides both coding gain and diversity advantage. In this dissertation, a new class of codes is proposed that extends the concept of turbo coding to include space-time encoders as constituent building blocks of turbo codes. The codes are referred to as turbo spacetime coded modulation (turbo-STCM). The motivation behind the turbo-STCM concept is to fuse the important properties of turbo and space-time codes into a unified design framework. A turbo-STCM encoder is proposed, which consists of two space-time codes in recursive systematic form concatenated in parallel. An iterative symbol-by-symbol maximum a posteriori algorithm operating in the log domain is developed for decoding turbo-STCM. The decoder employs two a posteriori probability (APP) computing modules concatenated in parallel; one module for each constituent code. The analysis of turbo-STCM is demonstrated through simulations and theoretical closed-form expressions. Simulation results are provided for 4-PSK and 8-PSK schemes over the Rayleigh block-fading channel. It is shown that the turbo-STCM scheme features full diversity and full coding rate. The significant gain can be obtained in performance over conventional space-time codes of similar complexity. The analytical union bound to the bit error probability is derived for turbo-STCM over the additive white Gaussian noise (AWGN) and the Rayleigh block-fading channels. The bound makes it possible to express the performance analysis of turbo-STCM in terms of the properties of the constituent space-time codes. The union bound is demonstrated for 4-PSK and 8-PSK turbo-STCM with two transmit antennas and one/two receive antennas. Information theoretic bounds such as Shannon capacity, cutoff rate, outage capacity and the Fano bound, are computed for multiantenna systems over the AWGN and fading channels. These bounds are subsequently used as benchmarks for demonstrating the performance of turbo-STCM.