Joint MCS Selection, Number of Stream and Stream Power Allocation for MIMO System in 5G

Adaptive Modulation and Coding (AMC) is a key technology to achieve high throughput in wireless communication. However, it only considers adaptive modulation and coding scheme (MCS) and ignores other factors which can also influence throughput. This paper focuses on the joint MCS selection, number of stream and stream power allocation for 5G system, with the target of maximizing average successful transmitted bits of the transport block. Two algorithms are proposed to solve it, which are Interior Point-Dichotomy (IPD) iterative algorithm and Power-based Deep Reinforcement Learning (PDRL) algorithm. The simulation results demonstrate the superiority of the two algorithms we propose. Specifically, IPD and PDRL algorithm respectively offer 0.04dB and 0.18dB gains with respect to adaptive MCS and number of stream (AMNS) on average power scheme.