This letter investigates the resource allocation problem for multiple Unmanned Aerial Vehicles (UAVs)-served Machine-to-Machine (M2M) communications. Our goal is to maximize the sum-rate of UAVs-served M2M communications by jointly considering the transmission power, transmission mode, frequency spectrum, relay selection and the trajectory of UAVs. In order to model the uncertainty of stochastic environments, we formulate the resource allocation problem to be a Markov game, which is the generalization of Markov Decision Process (MDP) for the case of multiple agents. However, owning to the UAVs mobility poses the difficulty of perceiving the environment, we propose a Long Short-Term Memory (LSTM) with Generative Adversarial Networks (GANs) framework to better track and forecast the UAVs mobility and improving the network reward. Numerical results demonstrate that the proposed framework outperforms the conventional LSTM and Deep Q-Network (DQN) algorithms.

In this letter, we investigate the resource allocation problem for cellular Zero-Touch Deterministic Industrial Machine-to-Machine (ZT-DI-M2M) communications. We consider a scenario, in which multiple self-adaptive industrial M2M communications are deployed underlaying cellular networks to emulate a variety of vertical use cases in Industry 4.0 with the objective of guaranteeing the determinacy of latency. Specifically, in order to figure out the relationship between latency determinacy and the utilization of wireless resources meanwhile modeling the uncertainty of the stochastic environment, we firstly formulate the resource allocation problem as a stochastic game by jointly considering transmission power control, frequency spectrum allocation and the selection of base stations. Afterwards, a random graph-based sparse Long Short-Term Memory (LSTM) network is proposed to solve the optimization problem while reducing the computational complexity. Finally, numerical result demonstrates the effectiveness of the proposed scheme.

Reliability is a critical factor in designing of Wireless Body Area Networks (WBANs). In this letter, we propose a resource scheduling strategy and solving an optimization problem to maximize the reliability of the transmission of emergency-critical sensory data. We jointly consider transmission mode, relay selection, time slot allocation and transmit power of each body sensor and formulating the scheduling problem to be a Markov Decision Process (MDP). In this strategy, the scheduling decision is made by each body sensor that do not have complete and global network information. Owning to the formulated problem is non-convex and the high computation complexity, we propose a deep reinforcement learning algorithm to solve the problem. Numerical results reveal that the proposed strategy is capacity of guaranteeing the reliability of transmission with an acceptable convergence speed.

As one of the most important metrics to sustainably provide communication services in green Industrial Internet of Things (IIoT), the problem of improving energy efficiency has constantly attracted extraordinary concerns from industry so far as to academia. In this paper, we intend to investigate the energy efficiency issue for an Energy Harvesting (EH)-enabled Machine-to-Machine (EH-M2M) communication underlaying Unmanned Aerial Vehicles (UAVs) networks from the perspective of resource management. Specifically, we are aiming at maximizing the average energy efficiency of EH-M2M communications by conjointly considering the EH time slot assignment, transmit power control and bandwidth allocation under the limitations of Quality of Service (QoS) and the available energy status of the EH-M2M devices. However, as the optimization problem is non-convex and NP-hard which is hard to tackle directly, we first transform the primitive objective function into a convex form equivalently by non-linear fractional programming and variable relaxation approach. After that, an iterative algorithm on the basis of Dinkelbach and Lagrangian theory is designed to optimize the resource management strategy. Finally, extensive simulation results demonstrate that the proposed scheme can establish more energy efficient communications compared to the benchmark schemes in different network settings.

Leaf diseases such as Mosaic disease and Black Rot are among the most common diseases affecting apple leaves, significantly reducing apple yield and quality. Detecting leaf diseases is crucial for the prevention and control of these conditions. In this paper, we propose incorporating rotated bounding boxes into deep learning-based detection, introducing the ProbIoU loss function to better quantify the difference between model predictions and real results in practice. Specifically, we integrated the Plant Village dataset with an on-site dataset of apple leaves from an orchard in Weifang City, Shandong Province, China. Additionally, data augmentation techniques were employed to expand the dataset and address the class imbalance issue. We utilized the EfficientNetV2 architecture with inverted residual structures (FusedMBConv and S-MBConv modules) in the backbone network to build sparse features using a top–down approach, minimizing information loss. The inclusion of the SimAM attention mechanism effectively captures both channel and spatial attention, expanding the receptive field and enhancing feature extraction. Furthermore, we introduced depth-wise separable convolution and the CAFM in the neck network to improve feature fusion capabilities. Finally, experimental results demonstrate that our model outperforms other detection models, achieving 93.3% mAP@0.5, 88.7% Precision, and 89.6% Recall. This approach provides a highly effective solution for the early detection of apple leaf diseases, with the potential to significantly improve disease management in apple orchards.

Semantic segmentation of remote sensing images is extensively used in crop cover and type analysis, and environmental monitoring. In the semantic segmentation of remote sensing images, owning to the specificity of remote sensing images, not only the local context is required, but also the global context information makes an important role in it. Inspired by the powerful global modelling capability of Swin Transformer, we propose the LSENet network, which follows the encoder-decoder architecture of the UNet network. In encoding phase, we propose spatial enhancement module (SEM), which helps Swin Transformer further enhance feature extraction by encoding spatial information. In decoding stage, we propose local enhancement module (LEM), which is embedded in the Swin Transformer to improve the Swin Transformer to assist the network to obtain more local semantic information so as to classify pixels more accurately, especially in the edge region, the adding of LEM enables to obtain smoother edges. The experimental results on the Vaihingen and Potsdam datasets demonstrate the effectiveness of our proposed method. Specifically, the mIoU metric is 78.58% on the Potsdam dataset, 72.59% on the Vaihingen dataset and 64.49% on the OpenEarthMap dataset.