Summary Nowadays, Ultra-wideband (UWB) technology is gaining much attention in the field of wireless communication, mainly due to its unique capabilities and potential applications together with the large unlicensed large frequency band devoted to it. However, the design and successful implementation of UWB transmission systems require a good understanding of the propagation characteristics of the radio channel. This thesis deals with the development of realistic channel models to support UWB communication design and evaluation. Different ways exist to develop a UWB channel model. A major method, which is the focus of this thesis, is to use statistical approaches based on measurements. To this end, a time-domain measurement set-up covering the frequency band from 3.1 to 10.6 GHz has been developed. Different limitations which are specific for the time-domain measurement are highlighted and solutions have been proposed to remove these limitations. As the performance of any communication system is determined by the characteristics of the radio channel in which it operates, our measurements have been conducted in different environments e.g. indoor office, indoor industrial area, outdoor to indoor and and Wireless Personal Area Network (WPAN) environments. In order to provide a good statistical UWB channel characterization, a sufficient number of channel impulse responses (CIRs) are gathered. These CIRs are retrieved from the time-domain measured signals using an appropriate post-processing. Based on the measured data, different statistical channel models, including large, middle and small-scale fading, have been proposed for the frequency band 3.1-10.6 GHz. UWB is a promising technology for several short-range applications (e.g. WPAN, WBAN). However, in these applications the transmit and receive antennas are very close to each other and far-field antenna conditions, assumed in most available link budget models, may not be satisfied. Under the near-field conditions, variations in the link budget and pulse shape compared to the far-field can be observed. In this thesis, a new empirical UWB link budget model for short-range UWB channels is proposed and validated by measurements and simulations using different types of antennas. The proposed model includes frequency, antenna size and orientation as parameters, and shows a good agreement with the simulations and the measurements. Usually in the statistical modeling of the UWB channel, the whole UWB band is sounded. Because of the very large bandwidth, the measurement set-up becomes complex, and consequently different parameters must be taken into consideration. In this thesis, a new approach to estimate the wireless UWB channel is introduced where partial information of sub-band channels is available (measured) and used in the estimation process. Positioning information in indoor environments is becoming more important and attractive, especially for wireless ad-hoc networks and area dependent wireless services. One of the main issues in positioning is how to achieve a good range estimation. Usually, the required range accuracy depends on the application and can vary from a few centimeters to tens of meters. UWB radio seems to be a promising solution for achieving a high accuracy. The last part of this thesis deals with the application of UWB technology for ranging/positioning. For instance, different methods based on time of arrival or signal strength, to estimate the range in dense indoor multipath environments are evaluated