A CDMA based hybrid wireless architecture for symmetric and asymmetric communications

by Wong, Wilson Wai

A CDMA Based Hybrid Wireless Architecture for Symmetric and Asymmetric Communications by Wilson Wai Shun Wong. Doctor of Philosophy, 2000. Graduate Department of Electrical and Cornputer Engineering, University of Toronto. The objective of this thesis is to study a code division multiple access (CDMA) based hybrid wireless architecture in which broadcast systems using singIe fiequency broadcasting are spectraily underlaid in an FDD-CDMA cellular network, for the provision of symmetric and asymmetric data services. A novel duplex scheme cailed Space Division Duplex (SDD) is proposed to support asymmetric communications. The use of a hybrid FDD/SDD-CDMA system architecture is then explored for supporting symmetric and asymmetric data users, with the former served by a cellular network and the latter served by the broadcast systems. The cellular network utilizes FDD to provide wide area voice and low-bit-rate symmetric data services, while the broadcast systems ut ilize SDD to provide high-bit-rate asymmetric data services. Using per-ce11 capacity as the performance measure, the capacity performance of this system architecture is evaluated and compared with two other classical systern architectures. The simulation results indicate that using adaptive array antenna structures at the broadcast system radio ports can improve capacity performance significantly and that more asymmetric data users can be supported at a given number of symmetric data users as the antenna beamwidth decreases. For example, when the number of symmetric data users is 30: the number of asymmetric data users increases hom 16 to 32 when the antenna beamwidth decreases Ekom 60° to 30°. Finally. an iterative SIR-based power control algorit hm is studied and its performance is accessed with the use of space diversity schemes. In particular, we have modeled the symmetric data users as voice users and the asyrnmetric data users as Internet users. When the voice user loading per sector is 80%: iterative power control with selection combining requires on average 12 access attempts for each Internet user compared to 18 access attempts with non-iterative power control. The best performance is achieved by using iterative power control with maximum ration combining, which requires on average 4 access attempts; and no sipificant degradation results until the voice user loading level is so high.