UC San Diego

The home and office environments are experiencing an ever increasing penetration of consumer electronic devices, often requiring data rates well in excess of tens of megabits per second. Communication to and from such devices has mostly relied on wireline technologies such as USB, DVI and IEEE1394. Ultra-wideband (UWB) and millimeter -wave (mmW) systems have been proposed to replace these wireline communication systems with short range high speed wireless networks. The significantly higher occupied bandwidth of UWB and mmW systems provides immense advantages in terms of higher data rates, while at the same time presenting new challenges such as greater susceptibility to interferers and possibly complex transceiver design. This dissertation addresses several technical challenges in the design of UWB and mmW systems. Multiple antenna techniques to improve the interference suppression capabilities and reliability of the UWB and mmW systems are employed. First, a MIMO beamforming system is analyzed. In the presence of antenna correlation and noisy channel estimates, an optimal MIMO beamforming scheme is proposed. The performance of this scheme is analyzed through a closed-form expression for the probability of error, and the combined effects of channel estimation errors and diversity on the system performance are studied. At the receiver, a fixed length antenna array is considered due to spatial constraints. For such an array, it is shown that there exists an optimal number of receive antenna elements for a given array length. The performance of a DS-CDMA-based UWB system with multiple antennas at the receiver is then analyzed. An optimal spatio-temporal receiver is proposed and its performance evaluated in the presence of narrowband interference, multiple access interference, antenna correlation and channel estimation errors. For a fixed array length and fixed maximum diversity level, the tradeoff between the number of antennas and the number of temporal taps in order to achieve the best performance is investigated. A 60 GHz mmW system is considered next. Multiple antenna equalization scheme to suppress both the intersymbol interference and multiple access interference is employed. A spectrally efficient multilevel quadrature amplitude modulation and a realistic IEEE channel model are used for analysis. The combined effect of interference suppression and spatial correlation on the system performance is studied through an analytically derived expression for bit error rate. It is shown that joint spatial and temporal processing can significantly improve the system performance