No description available.

Thesis (M.A.)--University of Hong Kong, 2006.

Thesis (M.S.)--Worcester Polytechnic Institute. / Keywords: vacuum; shape and deformation measurement; MEMS; vibrometry; scanning white light; Interferometry; thermal; vibration. Includes bibliographical references (136-139 leaves ).

Using a novel paradigm, we provide direct evidence for the role of inhibition during semantic interference resolution in memory. In Experiment 1, target words were primed via a naming task. Then, a cue word prompted participants to generate either a meaningfully related or unrelated word. Producing an unrelated word should require suppression of the cue's closest associates, which were the primed targets. Finally, participants read a list of words including the suppressed targets. Participants were slower to name targets in the unrelated condition than in the related condition, indicating that generating an unrelated word required suppression of competitors. Experiment 2 eliminated the initial priming phase, and a robust suppression effect was observed. Both studies showed that naming targets in the unrelated condition was even slower than controls. These results reflect that resolving semantic competition entails suppression of rejected competitors to below baseline levels, supporting an inhibitory account of interference resolution.

Using a novel paradigm, we provide direct evidence for the role of inhibition during semantic interference resolution in memory. In Experiment 1, target words were primed via a naming task. Then, a cue word prompted participants to generate either a meaningfully related or unrelated word. Producing an unrelated word should require suppression of the cue's closest associates, which were the primed targets. Finally, participants read a list of words including the suppressed targets. Participants were slower to name targets in the unrelated condition than in the related condition, indicating that generating an unrelated word required suppression of competitors. Experiment 2 eliminated the initial priming phase, and a robust suppression effect was observed. Both studies showed that naming targets in the unrelated condition was even slower than controls. These results reflect that resolving semantic competition entails suppression of rejected competitors to below baseline levels, supporting an inhibitory account of interference resolution.

This thesis utilizes pulse shaping in characterization of high efficient wavelength tunable terahertz radiation. The influence of multi-pulses generation under different tunable chirp and time delay are characterized by Cross-FROG measurement. A strong enhancement of total power is observed from spectrum. Furthermore, the total power of terahertz radiation is reducing in same multi-pulses spacing experiment as chirp is increasing and negative chirp have higher THz total power than positive chirp. Meanwhile, we also theoretically and experimentally study the possibility of an alternative approach of enhancing the THz power. This enhancement is attributing to constructive interference between two independently generated THz pulses. This also reveals the nature of quantum coherence and potential of high power THz radiation.

Thesis (Ph. D.)--University of Michigan, 1928. / "Reprinted from Physical Review ... vol. 34, no. 8, October 15, 1929.

Thesis (M. Phil.)--University of Hong Kong, 2002. / Includes bibliographical references.

published_or_final_version / Applied English Studies / Master / Master of Arts in Applied Linguistics

Interference creates a fundamental barrier in attempting to improve throughput in wireless networks, especially when multiple concurrent transmissions share the wireless medium. In recent years, significant progress has been made on characterizing the capacity limits of wireless networks under the premise of global and instantaneous channel state information at transmitter (CSIT). In practice, however, the acquisition of such instantaneous and global CSIT as a means toward cooperation is highly challenging due to the distributed nature of transmitters and dynamic wireless propagation environments. In many limited CSIT scenarios, the promising gains from interference management strategies using instantaneous and global CSIT disappear, often providing the same result as cases where there is no CSIT. Is it possible to obtain substantial performance gains with limited CSIT in wireless networks, given previous evidence that there is marginal or no gain over the case with no CSIT? To shed light on the answer to this question, in this dissertation, I present several achievable sum of degrees of freedom (sum-DoF) characterizations of wireless networks. The sum-DoF is a coarse sum-capacity approximation of the networks, deemphasizing noise effects. These characterizations rely on a set of proposed and existing interference management strategies that exploit limited CSIT. I begin with the classical multi-user multiple-input-single-output (MISO) broadcast channel with delayed CSIT and show how CSI feedback delays change sum-capacity scaling law by proposing an innovative interference alignment technique called space-time interference alignment. Next, I consider interference networks with distributed and delayed CSIT and show how to optimally use distributed and moderately-delayed CSIT to yield the same sum-DoF as instantaneous and global CSIT using the idea of distributed space-time interference alignment. I also consider a two-hop layered multiple-input-multiple-output (MIMO) interference channel, where I show that two cascaded interfering links can be decomposed into two independent parallel relay channels without using CSIT at source nodes through the proposed interference-free relaying technique. Then I go beyond one-way and layered to multi-way and fully-connected wireless networks where I characterize the achievable sum-DoF of networks where no CSIT is available at source nodes using the proposed space-time physical-layer network coding. Lastly, I characterize analytical expressions for the sum spectral efficiency in a large-scale single-input-multiple- output (SIMO) interference network where the spatial locations of nodes are modeled by means of stochastic geometry. I derive analytical expressions for the ergodic sum spectral efficiency and the scaling laws as functions of relevant system parameters depending on different channel knowledge assumptions at receivers. / text