Several emerging wireless applications require direct transmission between mobile terminals. Examples of these applications are mobile ad-hoc wireless networks, intelligent transportation systems, relay-based cellular networks, and future combat systems. Development of these mobile-to-mobile (M-to-M) systems depends on a good characterization of channel propagation. Another important consideration in modern communication systems is the use of multipath propagation to improve reliability and capacity of wireless systems. This is achieved by employing multiple antennas in multiple-input multiple-output (MIMO) systems and using techniques such as transmit and receive diversity. Considering the demand for high-speed wireless services, MIMO M-to-M systems are the leading candidates for future communication systems. To enable the successful design of MIMO M-to-M systems, our research focuses on modeling of MIMO M-to-M multipath fading channels and on diversity techniques for MIMO systems. Specifically, we propose two-dimensional (2-D) and three-dimensional (3-D) MIMO M-to-M statistical channel models that encompass narrowband and wideband MIMO channel scenarios for macro- and micro-cell environments. Furthermore, we validate the new models against measured data and find very close agreement between them. Using our 3-D models, we also investigate different antenna array configurations and their effect on the capacity of MIMO M-to-M systems. Contrary to common assumptions, we have found that there is no significant loss of capacity if the antenna array is tilted from the horizontal plane. Finally, we propose the design criteria for space-time coded continuous phase modulated systems. Our work would provide other researchers the tools needed to design and test future MIMO M-to-M communication systems.