Data Fusion With Communication Applications

Abstract : Data fusion ideas were employed to solve some important communication system problems. The first topic involves studying methods for generating and combining decisions from individual receivers in a multiple user scenario. The resulting theory of distributed multiuser detection or multiuser decision fusion has applications in both communications and in more classical problems of data fusion. In current digital cellular networks, for example, this theory is useful for combining decisions made at several remote base stations. This project focused on designing algorithms for and analyzing the performance and complexity of such schemes. The approach involved developing the theory for distributed multiuser detection by building on single user distributed detection theory, to which the PI has been a major contributor over the last few years. The second topic involved the study of methods employing multiple transmit and receive antennas (MIMO methods) to enhance performance. We developed new improved space-time codes and gave particular attention to the use of MIMO methods in a system (as opposed to a link which is interference free) which has not received much study.