Due to the enormous occupied bandwidth even by a single pair of UWB users, one major scientific challenge in UWB communications is interference management. Recently, Interference Alignment (IA) has become popular not only to well manage the interference, but also to optimally exploit the possible capacity gain caused by multiple pairs of transmitters and receivers. Theoretically, IA scales the channel capacity by K/2, where K is the number of user pairs. This fact makes IA highly attractive for future communication systems with numerous pairs of users. However, in the literature it has been reported that IA is not robust [7] against imperfections such as channel estimation errors. Thanks to the interdisciplinary research of antenna, communication and hardware engineers within the UKoLoS project Decimus, we were able to jointly search for solutions to improve IA robustness while not suffering from too perfect simulation idealities or too unrealistic hardware requirements. We find that antenna or pattern selection is a promising approach to improving the robustness of IA while keeping the underlying algorithms at a reasonable complexity and feasibility for implementation. Our contribution is structured as follows: first, a brief introduction of IA tailored to UWB is given. Then, we propose an antenna selection algorithm with low complexity and demonstrate its performance. In the third chapter a general methodology of MIMO UWB antenna design for orthogonal channels maximizing the channel capacity is presented. A first outcome of this methodology is a multi-mode orthogonal antenna which has been used for the investigated antenna selection approach. At last, the hardware requirements of IA systems with the proposed antenna selection method are studied. A conclusion summarizes the outcomes of our contribution.