Adaptive Modulation and Coding Based on Reinforcement Learning for 5G Networks

Abstract

We design a self-exploratory reinforcement learning (RL) framework, based on the Q-learning algorithm, that enables the base station (BS) to choose a suitable modulation and coding scheme (MCS) that maximizes the spectral efficiency while maintaining a low block error rate (BLER). In this framework, the BS chooses the MCS based on the channel quality indicator (CQI) reported by the user equipment (UE). A transmission is made with the chosen MCS and the results of this transmission are converted by the BS into rewards that the BS uses to learn the suitable mapping from CQI to MCS. Comparing with a conventional fixed look-up table and the outer loop link adaptation, the proposed framework achieves superior performance in terms of spectral efficiency and BLER.

Dynamic Bayesian Approach Applied in Link Adaptation Problem for Fifth Generation Network

“…In addition, a bandwidth with a frequency of 28 GHZ has been defined. We also consider an urban macro scenario, with a geometric channel model [19] and [20] given…”

“…, where λ is the transmission wave length and d is the antenna spacing. This channel model was developed based non-lineof-sight, in which shadowing was modeled by to a log-normal distribution with standard deviation of 6 dB, according [20]. The simulation starts with the UE moving away from the BS at a speed of 5km/h, with start and end points equal to 20m and 100m from the BS, respectively.…”

“…A physical layer model (PHY) in the 5G NR standard was adopted according to [20] and [21]. This standard supports quadrature phase shift keying (QPSK), and three types of modulation of quadrature amplitude modulation (16QAM, 64QAM and 256QAM) for Physical downlink shared channel (PDSCH).…”

Massive MIMO Adaptive Modulation and Coding Using Online Deep Learning

“…In the recent state-of-the-art there is a tendency to use Reinforcement Learning (Q-learning) technique for choosing the proper MCS like an agent action [10,11]. Our suggested approach works using the similar principle.…”

“…As a competitor of our solution we consider the following Q-Learning regression model [10,11] is true, and value 0 otherwise. The condition corresponds to the receipt of the success acknowledgement.…”