Search

Integrating the energy-harvesting (EH) relay and non-orthogonal multiple access (NOMA) technologies into narrow band internet of things (NB-IoT) networks can efficiently improve the energy and spectrum efficiency of the network and the quality-of-service of edge users. Therefore, we consider an EH relay aided NOMA NB-IoT network in this letter. To reduce the rate variance among NB-IoT devices, we aim to maximize the proportional fairness of data rate across all NB-IoT devices through jointly optimizing the communication resource allocation and successive interference cancellation (SIC) ordering subject to the minimum data rate requirements. Considering the non-convexity of this optimization problem, we propose a deep reinforcement learning based online optimization algorithm to obtain the sub-optimal solution. Simulation results demonstrate that the proposed algorithm can efficiently improve the proportional fairness and the total throughput among NB-IoT devices, in comparison with orthogonal multiple access techniques.

Deep learning-based algorithms provide a promising solution to efficiently generate offloading decisions in mobile edge computing (MEC) networks. However, considering dynamic MEC devices or offloading tasks, most of them require large-scale training data and long training time to retrain the deep neural networks (DNNs). In this letter, we propose a MEta-Learning-based computation Offloading (MELO) algorithm for dynamic computation tasks in MEC networks. Specifically, it learns from historical MEC task scenarios and adapts to a new MEC task scenario with a few training samples. Numerical results show that the proposed algorithm can adapt to a new MEC task scenario and achieve 99% accuracy via 1-step fine-tuning using only 10 training samples.

In this letter, we investigate the joint channel bandwidth and rate allocations for downlink non-orthogonal multiple access with the objective of maximizing the resource utilization efficiency (RUE). In particular, we take into account the utility of base station (BS) in serving the mobile terminals’ traffic and the cost in consuming channel bandwidth and measure the RUE as the ratio between the BS’s utility and the total power consumption. Despite the non-convexity of the joint optimization problem, we propose a three-layered vertical decomposition and design an efficient algorithm to compute the optimal solution. Extensive numerical results are provided to validate the performance of our proposed algorithms.

The recent small-cell dual-connectivity (DC) in 3GPP specification enables a mobile user (MU) to communicate with the macrocell and offload traffic through a small cell, which facilitates a flexible traffic offloading. The advanced non-orthogonal multiple access (NOMA) further enables small/macro cells to simultaneously accommodate multiple MUs’ offloading using the same channel, which improves spectrum-efficiency. In this letter, driven by the growing demand for greening cellular systems, we investigate the energy-efficient NOMA-enabled traffic offloading through DC, which aims to minimize the total power consumption of the small and macro cells, while satisfying each MU’s QoS-requirement. The optimal traffic offloading corresponds to a two-sided power allocation problem, which, however, is nonconvex. By exploiting the feature of DC, we equivalently transform this power allocation problem into a convex rate-splitting problem, which can be efficiently solved. Numerical results are provided to validate the advantage in saving power consumption by using the proposed optimal NOMA-enabled traffic offloading scheme.

In this letter, we propose an energy efficient power control scheme for resource sharing between cellular and device-to-device (D2D) users in cellular network assisted D2D communication. We take into account the circuit power consumption of the device-to-device user (DU) and aim at maximizing the DU's energy efficiency while guaranteeing the required throughputs of both the DU and the cellular user. Specifically, we define three different regions for the circuit power consumption of the DU and derive the optimal power control scheme for each region. Moreover, a distributed algorithm is proposed for implementation of the optimal power control scheme.

This letter studies the relay selection scheme for dual-hop decode-and-forward (DF) relay networks over Rayleigh fading channels. In particular, we first provide the problem formulation that minimizes the total transmit power consumption under the constraints of maximum tolerable outage probability and limited power when relay k cooperates. We then derive a closed-form expression of optimal power allocation that solves this optimization problem. Based on the derived closed-form expression, we can quickly select the best relay that consumes the minimum power during the cooperation of transmission.

In this letter, we first prove that the upper boundary of the feasible signal to interference-plus-noise ratio (SINR) region achieved by power control is Pareto optimal under the constraints of minimum SINR requirements and limited power. Thanks to the Pareto optimal nature, we further propose a bisection searching based low-complexity algorithm that achieves arbitrary Pareto optimal power control over the entire feasible SINR region. More importantly, we can easily approach the upper boundary of the feasible SINR region through the proposed algorithm.