Accurate prediction of user traffic in cellular networks is crucial to improve the system performance in terms of energy efficiency and resource utilization. However, existing work mainly considers the temporal traffic correlation within each cell while neglecting the spatial correlation across neighboring cells. In this letter, machine learning models that jointly explore the spatio-temporal correlations are proposed. Specifically, several recurrent neural network structures are utilized. Furthermore, a multi-task learning approach is adopted to explore the commonalities and differences across cells in improving the prediction performance. Base on real data, we demonstrate the benefits of joint learning over spatial and temporal dimensions.

In this paper, we study a joint placement, resource allocation, and user association problem for UAV-assisted wireless networks with constrained backhaul links, where multiple UAV-mounted base stations (UBSs) are deployed to provide wireless services for ground users. We propose a novel framework to maximize the user throughput within the flight-time of UBSs and provides fairness among the users. We first obtain the optimal resource allocation schemes based on different fronthaul and backhaul conditions, and an efficient iterative algorithm is then developed to jointly optimize user association and UBS placement. The optimal UBS placement can be achieved by solving an unconstrained optimization problem which is a simplification of the initial constrained optimization problem based on the optimal resource allocation. We develop a dual-domain coordinated descent and bipartite graph matching based sub-process to identify an optimal user association that prefers the nearby UBSs, as the user association under constrained backhaul links have non-unique optimal solutions. Extensive simulations are conducted to verify the effectiveness of the proposed algorithm, and results show that our proposed method under constrained backhaul can improve both the average throughput by 49% and the fairness among the users by 47% in comparison with the method under ideal backhaul.

Distributed power allocation is important for interference-limited wireless networks with dense transceiver pairs. In this paper, we aim to design low signaling overhead distributed power allocation schemes by using graph neural networks (GNNs), which are scalable to the number of wireless links. We first apply the message passing neural network (MPNN), a unified framework of GNN, to solve the problem. We show that the signaling overhead grows quadratically as the network size increases. Inspired from the over-the-air computation (AirComp), we then propose an Air-MPNN framework, where the messages from neighboring nodes are represented by the transmit power of pilots and can be aggregated efficiently by evaluating the total interference power. The signaling overhead of Air-MPNN grows linearly as the network size increases, and we prove that Air-MPNN is permutation invariant. To further reduce the signaling overhead, we propose the Air message passing recurrent neural network (Air-MPRNN), where each node utilizes the graph embedding and local state in the previous frame to update the graph embedding in the current frame. Since existing communication systems send a pilot during each frame, Air-MPRNN can be integrated into the existing standards by adjusting pilot power. Simulation results validate the scalability of the proposed frameworks, and show that they outperform the existing power allocation algorithms in terms of sum-rate for various system parameters.

The in-band wireless backhaul that shares spectrum resources between fronthaul and backhaul is a promising method for UAV-mounted base stations (UBSs) to provide wireless services with limited resources. However, the existing works mainly considered the out-band wireless backhaul, where the resources of fronthaul and backhaul are allocated separately. Moreover, the UBS placement methods usually focused on the optimization of fronthaul while neglecting the backhaul. In this paper, a placement algorithm that jointly optimizes the resource allocation and user association for multiple UBSs with in-band wireless backhaul is proposed. In particular, the joint optimal resource allocation for both fronthaul and backhaul is derived using convex optimization theory. Given the optimal resource allocation, an equivalent spectrum efficiency is defined to simplify the optimization problem. As a result, an effective algorithm is proposed by solving the user association problem and UBS placement problem iteratively. In addition, simulation results are provided to verify the efficiency and effectiveness of our proposed method.

The joint communication-radar system (JCRS) is promising in the future wireless network which can provide great advantages, such as structure simplification, interference mitigation and less resource occupation, compared with the traditional separate deployment of communication and radar systems. However, with the large number of users accessing the network both for communication and radar purposes, massive user data packets may arrive at the base station equipped with the JCRS in a short period of time, which can easily cause the backlog of user data packets and affect network stability. To solve this problem, we introduce orthogonal frequency division multiple access (OFDMA) technique into the JCRS to support communication function for massive users. Then, we design a queue scheduling model of user data packets to perceive the status of current queue backlog and schedule the queue dynamically. Different from the existing works, we formulate a transmit power minimization problem, while taking network stability and radar detection performance into account. Since the problem is a non-convex problem, we transform it into the upper-bound minimization of drift-plus-function by using the Lyapunov optimization technique. Further, the successive convex approximation method and parameter transformation method are exploited to convert this problem into a convex problem. Finally, we derive an optimal closed-form power allocation expression by using the Lagrangian dual decomposition method and propose a dynamic subcarrier and power allocation algorithm. Simulation results show that the proposed algorithm can achieve an [O(1/?),O(?)] power-backlog tradeoff and demonstrate the efficacy of the proposed algorithm in reducing transmit power while significantly enhancing network stability.

Due to the high mobility and easy deployment, unmanned aerial vehicles (UAVs) have attracted much attention in the field of wireless communication and positioning. To meet the challenges of lack of infrastructure coverage, uncertain sensor position and large amount of sensing data collection in wireless sensor network (WSN), this paper presents an efficient joint data collection and sensor positioning scheme for WSN supported by multiple UAVs. Specifically, a UAV is set as the main UAV to collect data, and other UAVs are used as auxiliary UAVs for sensor positioning using time difference of arrival (TDoA). A mixed-integer non-convex optimization problem with uncertain sensor position is established. The goal is to minimize the average positioning error of all sensors by jointly optimizing the UAV trajectories, sensor transmission schedule and positioning observation points (POPs). To solve this optimization model, the original problem is decomposed into two sub-problems based on the path discrete method. Firstly, the block coordinate descent (BCD) and successive convex approximation (SCA) techniques are applied to iteratively optimize the trajectory of the main UAV and the sensor transmission schedule, so as to maximize the minimum amount of data uploaded by the sensor. Then, based on the trajectory of the main UAV, a particle swarm optimization (PSO)-based algorithm is designed to optimize the POPs of UAVs. Finally, the spline curve is applied to generate the trajectories of auxiliary UAVs. The simulation results show that the proposed scheme can meet the requirements of data collection and has a good positioning performance.