Machine-to-Machine (M2M) communication plays a pivotal role in the evolution of Internet of Things (IoT). Cellular networks are considered to be a key enabler for M2M communications, which are originally designed mainly for Human-to-Human (H2H) communications. The introduction of M2M users will cause a series of problems to traditional H2H users, i.e., interference between various traffic. Resource allocation is an effective solution to these problems. In this paper, we consider a shared resource block (RB) and power allocation in an H2H/M2M coexistence scenario, where M2M users are subdivided into delay-tolerant and delay-sensitive types. We first model the RB-power allocation problem as maximization of capacity under Quality-of-Service (QoS) constraints of different types of traffic. Then, a learning framework is introduced, wherein a complex agent is built from simpler subagents, which provides the basis for distributed deployment scheme. Further, we proposed distributed Q-learning based autonomous RB-power allocation algorithm (DQ-ARPA), which enables the machine type network gateways (MTCG) as agents to learn the wireless environment and choose the RB-power autonomously to maximize M2M pairs' capacity while ensuring the QoS requirements of critical services. Simulation results indicates that with an appropriate reward design, our proposed scheme succeeds in reducing the impact of delay-tolerant machine type users on critical services in terms of SINR thresholds and outage ratios.

In this paper, we introduce a distributed power allocation strategy for random access, that has the capabilities of multipacket reception (MPR) and successive interference cancellation (SIC). The proposed random access scheme is suitable for machine-to-machine (M2M) communication application in fifth-generation (5G) cellular networks. A previous study optimized the probability distribution for discrete transmission power levels, with implicit limitations on the successful decoding of at most two packets from a single collision. We formulate the optimization problem for the general case, where a base station can decode multiple packets from a single collision, and this depends only on the signal-to-interference-plus-noise ratio (SINR). We also propose a feasible suboptimal iterative per-level optimization process; we do this by introducing relationships among the different discrete power levels. Compared with the conventional power allocation scheme with MPR and SIC, our method significantly improves the system throughput; this is confirmed by computer simulations.