Interferometric phase data acquired using a pair of synthetic aperture radar (SAR) sensors, need to be phase unwrapped before useful information can be extracted. This proves to be difficult because measurements tend to be corrupted by additive random phase noise. A common way to reduce random phase noise before phase unwrapping, is to make use of a coherent averaging or mean filter. Unfortunately, noise reduction is achieved at the expense of a loss in spatial resolution. We propose adopting the technique of using local statistics, to adaptively suppress the noise in the wrapped interferometric image. This technique has been applied to multiplicative SAR speckle noise filtering and shown to have good feature retention capabilities. However, because the local statistics (LS) filter is essentially designed for non-directional (noncircular) data, modifications have to be made before it can be applied in the wrapped-phase domain. For this purpose, an angular complement compensation technique is devised and circular variances are used. In instances when the local statistics of the signal can no longer be assumed stationary i.e. dense fringes with spacing in the order of the window size, the non-stationary phase signal is first removed using a two-dimensional multiple-signal classification (MUSIC) algorithm. The result is a noise-filtered wrapped interferogram with retained spatial resolution, independent of the density of the fringes. Quantitative evaluation using synthetic fringe patterns and computed residues, reveals that the noise reduction capabilities of the proposed LS filter is comparable to the mean filter.

This thesis describes the design of power-scalable digital adaptive equalizer for pulse or quadrature amplitude modulation communication systems, using synthesis and place-and-route tools. DSP based modem applications such as gigabit Ethernet transceivers require channel equalization. Because of the high rate and computation complexity involved, adaptive equalization filters consume a lot of power. Currently, equalization is typically hardwired instead of using a digital signal processor. Yet, there is a need for the equalization filters to be scalable to different channel and bit rate requirements. Synthesis and place-and-route tools enables the designer to focus on higher-level aspects of the design instead of at the transistor level. In this thesis, we have used adaptive tap length and precision techniques to design a digital adaptive equalizer whose power consumption is scalable to the precision requirements.