Power Allocation for Practicable Capacity Maximization in Eigen-MIMO

Water-filled eigenchannels offer the highest MIMO information-theoretic capacity. However, in practice there are many factors, such as digital modulations instead of Gaussian signals, finite block lengths, and imperfect power allocation, that combine to degrade the capacity from the Shannon limit to the practicable capacity - the uncoded throughput - of a digital link. Complexity reduction is also an important factor for a practical system. In this paper, we address the problem of power allocation for maximizing the practicable capacity in eigen-MIMO with the total input power constraint and perfect channel state information (CSI) at the transmitter. We use non-adaptive modulation at the transmitter in order to avoid the high complexity in both the electronics and the protocol - of adaptive modulation. The optimum power allocation reveals new and interesting capacity behavior, and it turns out to be different than that of the water-filling.