Adaptive MIMO RF systems: Post-manufacture and real-time tuning for performance maximization and power minimization

MIMO RF systems are seeing widespread use due to their inherent ability to extend the range of wireless channel conditions under which data can be transmitted reliably while satisfying communication throughput requirements. However, as frequencies scale upwards and aggressive CMOS technologies are used to manufacture RF devices, semiconductor process variations will limit the range of operating conditions under which MIMO systems can operate reliably with low power usage. To resolve this, we propose post-manufacture testing and tuning algorithms for MIMO systems assembled from devices across diverse process corners. These algorithms allow such systems to operate across the widest range of channel conditions as possible while minimizing power consumption (performance maximization also maximizes manufacturing yield). To further save power consumption in the field, real-time tuning algorithms are proposed that dynamically trade off available performance slack in MIMO designs, across the relevant signal modulation and MIMO modes to save power consumption. With regard to post-manufacture tuning, it is seen that as much as 20% yield improvement is possible through use of the proposed tuning methodology. With regard to real-time tuning, it is seen that as much as 3X power savings is possible for specific wireless channel conditions. The results of various experiments on MIMO systems are discussed.