Jiangsu Second Normal University

We, the Publisher, have retracted the following article: Yang-yang Hou (2024). Applications of Image Segmentation Techniques in Medical Images. https://doi.org/10.4108/eetel.4449 The article has been retracted due to misconduct during the peer review process. This does not implicate any wrongdoing from the author’s side. We informed the authors about this decision. The retracted article will remain, and it has been watermarked as “RETRACTED”.

This review systematically examines major research advances in AUV path planning over recent years, covering several mainstream methodologies: sample-based path planning (e.g., PRM and RRT along with their asymptotically optimal variants, suitable for high-dimensional space exploration), graph-search-based path planning (e.g., A-series and D-series algorithms, achieving global optimization and dynamic replanning through environmental modeling), optimization-based approaches (including artificial potential field (APF), nonlinear programming (NLP), and model predictive control (MPC), designed to satisfy stringent dynamic constraints on AUV motion), swarm intelligence-based planning methods (such as genetic algorithms and ant colony optimization), and learning-based intelligent methods (such as deep reinforcement learning (DRL) for real-time decision-making in unknown and dynamic environments). Through in-depth analysis of these methods’ principles, improvement strategies, and AUV path planning contexts, this review highlights current research trends toward hybrid cooperative planning, dynamic environmental adaptability, and high-precision trajectory optimization. Finally, the paper outlines future directions for AUV path planning, emphasizing multi-AUV collaboration and higher-level intelligent decision-making as key research priorities.

The gibbon was an often-mentioned animal in the elite literati world of China, Japan, and the Korean Peninsula during the twelfth to nineteenth centuries, with a large number of poems and paintings touching on the theme of gibbons. The image of the gibbon is often entangled with religion (Daoism and Zen Buddhism), and through the coding of art, it has become a tool for displaying one’s political position. There is also an intertextual phenomenon in the presentation of gibbons from poetry to painting which plays a crucial role in the generation of meaning. This article attempts to answer two questions: (1) why and how the image of a gibbon has become a tool for displaying religious affiliation and one’s political position through the art form of Sinitic poetry; (2) how does the intertextual relationship between poetry and painting affect the creation of meaning?

Hydrological time series forecasting faces significant challenges in balancing prediction accuracy and timeliness due to the inherent nonlinearity and non-stationarity of water level data. While Long Short-Term Memory (LSTM) networks are widely employed for their capability to capture long-term dependencies, they frequently exhibit prediction delays, particularly in complex hydrological scenarios. To address this limitation, we propose a hybrid model, CS-VMDLSTM, which integrates Variational Mode Decomposition (VMD) with Cuckoo Search (CS) optimization. The core innovation involves the adaptive decomposition of water level sequences into multiscale intrinsic mode functions (IMFs) via VMD, where its critical parameters (number of modes <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$K$</tex>, penalty factor <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\alpha$</tex>) are optimized using the CS algorithm. Each IMF subcomponent is subsequently processed by a dedicated LSTM sub-model. The final prediction is obtained by aggregating the outputs of all sub-models. Experimental evaluation on the Tunxi Basin dataset (2016-2018) demonstrates that the proposed model reduces prediction delays by an average of 1 hour across 2-6 hour lead times and improves the Nash-Sutcliffe Efficiency (NSE) by 7.6 % compared to the baseline LSTM. The CS-VMD-LSTM framework effectively mitigates peak lagging and enhances temporal responsiveness, providing a robust solution for real-time flood forecasting and water resource management.

As an emerging technology, digital twin (DT) can provide real-time status and dynamic topology mapping for Internet of Things (IoT) devices. However, the implementation of DT within IoT networks necessitates substantial, distributed data support, which raises privacy concerns. To address these issues, we develop a dynamic resource scheduling algorithm tailored for the asynchronous federated learning (FL)-based lightweight DT empowered IoT network. Specifically, our approach aims to minimize a multi-objective function that considers both energy consumption and latency by optimizing IoT device selection and transmit power control, while adhering to long-term performance constraints derived from the convergence upper bound for asynchronous FL. We utilize the Lyapunov method to decompose the formulated problem into a series of one-slot optimization problems and design a two-stage optimization algorithm to achieve optimal transmit power control and efficient IoT device scheduling. In the first stage, we derive closed-form solutions for optimal transmit power on the IoT device side. In the second stage, to address the uncertainty in partial state information, the edge server leverages a multi-armed bandit (MAB) framework to model the IoT device selection problem and employs an efficient online algorithm, the client utility-based upper confidence bound (CU-UCB), to solve it. Moreover, we analyze the algorithm’s optimality gap and time-averaged regret, further demonstrating its effectiveness in resource allocation. Numerical results confirm the superior performance of our algorithm compared to benchmark methods, with the CU-UCB algorithm achieving an 18.9% reduction in average cost and a 6% improvement in accuracy on real-world datasets.

Nowadays the flourishing development of communication, network, and sensor technologies has greatly empowered a wide variety of Internet of Things (IoT) applications. However, security and privacy concerns are invoked as tremendous unauthorized sensors and devices are connected with the IoT networks. Nonetheless, traditional resource-consuming authentication methods such as Public Key Infrastructure (PKI) cannot be deployed due to the limited computational and storage capacity of lightweight IoT nodes. When an authentication scheme is aided by blockchain, the trustworthiness and reliability of the authentication results can be improved and distributed computational and storage resources can be collected for authentication. Driven by these issues, we propose a blockchain-aided decentralized authentication framework over covert channels for lightweight IoT networks, named CovertBC. Specifically, the proposed framework orchestrates a physical-layer covert channel design for lightweight identity information transmission and a blockchain-layer consensus mechanism design for trusted and decentralized authentication. First, we develop an off-chain reputation management scheme to evaluate the reputation of the CovertBC nodes according to its correctness on decoding the received covert messages. Second, we propose a reputation-based practical Byzantine fault tolerance consensus protocol, in which the blockchain nodes with high reputation are selected to cooperatively authenticate the accessing node on the chain. Numerical results reveal that CovertBC can achieve higher authentication efficiency than the baseline schemes.

The severe energy consumption problem of 5 G base stations (BSs) has always been an urgent issue to be solved. Many BS energy-saving strategies have been proposed, the most classic of which is the reinforcement learning (RL) scheme. However, training an RL model requires online exploration of the optimal path. Since the switching on and off of BSs directly affects the local wireless network, the exploration cost is unacceptable. Subsequently proposed offline reinforcement learning schemes, such as Decision Transformer (DT), are only applicable to decisions with a time length shorter than the training samples and cannot be applied in practice. To address the issue of limited decision length in DTs, this paper proposes a Finite-State-Machine Decision Transformer (FSM-DT) for BS energy-saving. We remodel the energy-saving decision process as a finite-state machine cyclic process and design a new method for calculating the returns-to-go feature. This method can optimize the performance of the DT algorithm in infinite-horizon tasks during actual deployment. We also design an intermittent mask for the returns-to-go feature to avoid invalid attention. We conduct a series of experiments in multiple simulated environments. The results show that FSM-DT outperforms the DT model in decision-making performance and generalization ability, achieving a 19.3 % improvement in energy-saving performance and addressing the issue of unstable long-term decision-making in the DT model.

Indoor navigation for unmanned vehicles presents significant challenges due to confined spaces and dynamic obstacles. This study comparatively investigates Proximal Policy Optimization (PPO) and Advantage Actor-Critic (A2C) for dynamic path planning. We construct a simulation platform based on ROS and Gazebo with an enclosed environment, featuring a dynamically moving cylindrical obstacle that emulates human walking patterns. Both PPO and A2C are trained and evaluated under identical settings, enabling a comparative analysis of their stability, efficiency, and obstacle-avoidance performance. Results show that PPO consistently achieves higher success rates and smoother trajectories with fewer collisions.