| Wireless cellular communications has been one of the fastest growing fields of technology in the world. As opposed to its wireline counterparts, wireless communications poses some unique challenges including multipath fading and co-channel interference. Diversity techniques are overwhelmingly used in wireless communication systems to enhance capacity, coverage and quality, among which space diversity, i.e., diversity realized in space with antenna arrays, is favored because it does not impose a penalty in terms of scarce spectrum resources. Under the framework of wireless cellular communications with antenna arrays, both signal processing and information theoretic aspects are studied in this dissertation.; The signal processing techniques investigated are, among others, space-time processing, multiuser detection, and turbo decoding. All of these techniques exhibit near-Shannon-limit performance with reasonable complexities in many cases, and are very promising for next-generation communications. Specifically, various transmit diversity and downlink beamforming techniques with power control are examined and compared for wireless cellular communications with transmit arrays, and a range of iterative space-time multiuser detection techniques are explored with receive arrays. Further, turbo space-time multiuser detection techniques are employed for wireless cellular multiple-input multiple-output (MIMO) communications, i.e., with antenna arrays on both transmit and receive ends. Then, for multicell MIMO systems where co-channel interference is the dominating detrimental factor, various multiuser receivers are proposed to dramatically improve the system performance.; Spectral efficiency of MIMO systems operating in multicell frequency-flat fading environments is also studied. The following detectors are analyzed: a single-cell detector, the joint optimum detector, a group linear minimum-mean-square-error (MMSE) detector, a group MMSE successive cancellation detector, and an adaptive multiuser detector. Large-system asymptotic (non-random) expressions for these spectral efficiencies are developed. Some analytical and numerical results are derived based on these expressions to gain insight into the behavior of multicell MIMO systems.; Even though wireless cellular communications constitutes the main part of this dissertation, an application of some of the methods developed to wireline communications is also considered. In particular, the turbo multiuser detection techniques are applied to digital subscriber line (DSL) wireline communications to effectively combat crosstalk, with the influence of impulse noise taken into consideration. |
