This paper presents an analog implementation of a fast controller for a unity-power-factor (UPF) PWM rectifier. The best settling times of many popular controllers for this type of converter are on the order of a few line cycles, corresponding to bandwidths under 20 Hz [l]. The fast controller demonstrated in this paper can exercise control action at a rate comparable to the switching frequency rather than the line frequency. In order to accomplish this while maintaining unity power factor during steady-state operation, the fast controller employs a ripple-feedback cancellation scheme.

Unlike the slow control schemes associated with state-of-the-art UPF converters, where disturbances are attended to by control action taken only once per line cycle, the fast controller studied in this thesis makes use of the fact that control action can be exercised at a much faster rate (the switching frequency). As a result the fast controller, which is based on a large-signal periodically-varying model of a UPF power converter is able to achieve a much faster response time to disturbances. In addition, the fast controller may facilitate system improvements over conventional slow controllers such as the reduction of the output bus capacitance. A comparison is done between the conventional slow controllers and the fast controller. The robustness, stability and practical feasibility of the fast controller are examined. An analog implementation of a slow controller based on a large-signal model and the fast controller are presented.

ABSTRACT

This paper presents a technique for obtaining a well-conditioned channel matrix in a line of sight multiple input multiple output (MIMO) environment. The technique is based on the implementation of a back-to-back antenna system as a passive repeater to enhance performance in MIMO systems. The flexible configuration with no need for a phase controller allows to spread the proposed repeater in MIMO communications to ensure spatial multiplexing and enhance capacity. A condition number and matrix rank are proposed as metrics to demonstrate the validity of the proposed method.

In MIMO communications, high buildings in cities and other obstructions can prevent the propagation of radio waves, resulting in blind spots and poor communications. This problem can be solved using passive repeaters, which, by distributing them in the communication environment, the blind spots can be accessed and the coverage area can be extended. The distribution of these repeaters in the MIMO communication should be optimized in order to maximally cover the target area. In this article, an approach based on non-negative least square (NNLS) algorithm is proposed for optimizing the distribution of the minimum number of repeaters in none-line-of sight (NLOS) MIMO communications. The proposed approach is implemented using passive repeaters consisting of two parabolic antennas connected through a flexible cable. The numerical analysis is performed to verify the validity of the proposed approach, and it is found that the proposed method helps to optimally distribute passive repeaters and extend coverage area with minimum number of repeaters.