Ph.D Thesis | |

Ph.D Student | Zaidel Benjamin |
---|---|

Subject | Information Theoretic Aspects of CDMA Systems |

Department | Department of Electrical and Computer Engineering |

Supervisor | PROF. Shlomo Shamai )Shitz( |

In this thesis we investigate several information theoretic aspects of multicell communications. We start by focusing on the uplink and analyze system performance, under various assumptions, when optimally coded randomly spread direct sequence code-division multiple access (DS-CDMA) is employed. The setting adheres to

Wyner's (1994) infinite linear cell-array model, where only adjacent-cell interference is present, and characterized by a single parameter 0 £ a £ 1. The focus is on asymptotic analysis where both the number of users per cell and the processing gain go to infinity, while their ratio (the “cell load”) goes to some finite constant.

First, single cell-site processing is considered, and four
multiuser detection strategies, differing in the information on multiuser
interference utilized by the receiver, are analyzed and compared. The spectral
efficiency, specifying the ultimate performance, is evaluated for non-fading
channels, as well as for *slow* flat-fading, while employing symbol-level
interleaving. For the latter setting, a suboptimum practically oriented
transmission and decoding strategy is also considered, assuming that all users
employ equal rates and transmit powers, while considering a *strongest-users-only
decoding* scheme. The total capacities under an outage constraint, derived
as functions of the fraction of users that *cannot* be decoded, are
analyzed and compared to the corresponding spectral efficiency results. We then
proceed with the analysis the optimum and linear MMSE *joint multicell
receivers*, while assuming flat-fading and *chip-level interleaving*.
Dramatic performance enhancement as compared to *single*-cell-site
processing is demonstrated. The impact of chip-interleaving vs.
symbol-interleaving is also investigated. The final part of the thesis
investigates the average per-cell sum-rate capacity with joint multicell
processing, and* no random spreading*. The uplink, *downlink*, and
both non-fading and flat-fading channels are addressed. The focus is on a
simple Wyner-like model, where the system cells are arranged on a circle,
assuming the cell-sites are located at the *boundaries* of the cells.
Assuming *individual per-cell power constraints*, the sum-rate capacity of
the downlink is shown to coincide with the corresponding explicitly obtained
uplink result for non-fading channels. Introducing flat-fading, lower and upper
bounds on the sum-rate capacity are derived, exhibiting an *O*(log_{e}
*K*) multi-user diversity factor for a number of users per-cell *K *>>
1, in addition to the array diversity gain. Joint multi-cell processing is
shown to eliminate out-of-cell interference, which is traditionally considered
to be a limiting factor in high-rate reliable communications.