Abstract Due to the challenges posed by background noise and the limited information available for small targets in remote sensing images, the detection performance for such targets remains unsatisfactory. To address these issues and enhance detection accuracy, we propose an improved algorithm based on RTDETR, named Adaptive Selective Transformer. Firstly, in the feature extraction network, we introduce an adaptive convolutional feature enhancement module to improve the multi-scale feature extraction capability in low-resolution remote sensing images. Secondly, we design a multi-scale enhancement structure to extract detailed information from small target images through enhanced multi-scale representation learning, thereby generating target features with stronger discriminative power. Finally, we propose a hierarchical frequency attention mechanism to achieve localized enhancement of contextual awareness, effectively capturing high-frequency local feature information of small targets. Experimental results demonstrate that the Adaptive Selective Transformer achieves superior small target detection performance, validating the effectiveness of our modifications to the original RTDETR model.

Abstract Click-through rate (CTR) prediction has become a crucial task in online advertising and other fields. Many researchers focus on improving CTR prediction models by exploring feature interactions. One popular model, Deep Factorization Machine (DeepFM), addresses both high-order and low-order feature interactions, but it overlooks the variability of feature representation in different contexts and lacks a comprehensive explanation of high-order feature interactions. In this paper, we propose a CTR prediction model called DeepFM-GA, which is based on improved feature refinement generation and attention enhancement representation. Firstly, we incorporate an attention convolutional generation module into $\text{DeepFM}_{\text{FRNet}}$, which enriches the feature space by generating complementary features through convolutional neural networks while maintaining context-aware feature representation. Secondly, we utilize a multi-head self-attention layer for feature-enhanced representation, enhancing the model’s ability to select important features. Finally, experiments are conducted on four real-world datasets, and the results show that DeepFM-GA has a better performance compared to other mainstream CTR models.

ABSTRACT In this study, a traffic flow prediction framework based on multigraph structural modeling and dynamic sensing of key nodes is proposed to address the significant differences in traffic patterns between holidays and workdays. Firstly, the original data are categorized using calendar labels to generate holiday‐indication tags for each sample. Then, a sliding‐window sample entropy method is employed to identify highly dynamic key nodes from the workday and holiday datasets to construct the key‐node set. A dynamic masking mechanism based on the calendar labels is introduced during model training to enhance the transformer's attention to core areas. To fully capture the connectivity variations of the traffic system under different contexts, this study adopts a dual adjacency matrix design and dynamically switches the graph structure based on sample labels during forward propagation, enabling context‐aware spatial modeling. Finally, traffic prediction accuracy is effectively improved by integrating a graph convolutional network (GCN) to extract spatial features with an enhanced transformer to dynamically model temporal dependencies. Experimental results demonstrate that this method significantly outperforms traditional single‐graph structure models across various calendar scenarios, validating the effectiveness and potential of multigraph dynamic modeling with a key‐node mechanism for traffic prediction tasks. In addition, ablation experiments are also conducted to compare the effects of different key‐node selection and processing strategies.

Vehicle ad hoc network (VANET), with its characteristics of fast mobility and uneven distribution, adds complexity and uncertainty to the network. To ensure reliable routing connectivity in VANET, it is crucial to tackle challenges, such as implementing small-scale and low-precision solutions based on the spanning tree, as well as addressing the absence of effective contingency plans in case of failures. Constructing a large-scale suboptimal spanning tree solution set (LST) becomes the key to solving the aforementioned problems. Previous researchers have utilized swarm intelligence optimization algorithms to address the spanning tree construction problem. However, these methods suffer from drawbacks, such as low precision, poor scalability, lack of diversity, and uneven distribution. To tackle the aforementioned issues, this article proposes a quantum binary artificial bee colony algorithm (QBABC). First, the spanning tree hit ratio (SHR) is introduced to evaluate the probability of acquiring a spanning tree in VANET. Second, a mathematical model and data set are constructed based on the considered Quality-of-Service (QoS) metrics. Then, a quantum random number generator (QRNG) is proposed, which incorporates a fusion of binary encoding strategies. Finally, a multistage search strategy inspired by honeybee behavior is adopted. Through nonparametric statistics and validation with corresponding metrics, the results demonstrate that QBABC exhibits strong competitiveness and provides effective solutions in the event of VANET failures. QBABC’s advantages lie in improving the precision, scalability, diversity, and uniformity of spanning tree construction. This research is of significant importance for enhancing reliable routing connectivity in VANET.

With the advancement of automatic driving and smart city, it is critical to predict traffic information for traffic management, traffic planning, and traffic safety. When predicting traffic information, the spatial structure of the roads will also affect the traffic flow information, such as speed, occupancy rate, etc. The common method either merely focusing on the temporal feature without the considering the spatial structure, or the method of spatial feature extraction is only applicable to Euclidean structure, which does not apply to Non-Euclidean structure. This paper proposes a traffic speed prediction method based on time classification in combination with spatial Graph Convolutional Network. This method employs Gated Recurrent Unit to extract the temporal correlation and Graph Convolutional Network to extract the traffic network’s spatial structure. In consideration of the varying features of traffic speed on weekdays and weekends in the time dimension, time is divided into two types: weekdays and weekends. Since the structure of the road network will not change in the short term in actual process, the same network structure of spatial graph convolution can reasonably be shared in the spatial dimension after which the two sections are fused for training and prediction. Finally, this proposed method is compared to some baseline models to prove the performance. Generally speaking, this strategy produces more accurate prediction results on the PEMS_BAY and METR_LA data sets than the baseline models.

With the rapid development of Internet of Things (IoT) in the field of transportation, the vehicle-to-vehicle (V2V) communication not only becomes available on a large scale, but also will be an indispensable part of the future transportation. License plates are the identification of vehicles, so the license plate detection and recognition in the V2V communication scenario is very important. However, the existing license plate detection and recognition methods are suffering from a low accuracy rate issue. To solve this issue, we propose a hybrid deep learning algorithm as the license plate detection and recognition model by fusing YOLOV3 and CRNN. The proposed model enables the network itself to better utilize the different fine-grained features in the high and low layers to carry out multi-scale detection and recognition. In this model, we utilize the fast and accurate performance of YOLOV3, and the excellent detection ability of CRNN. As a result, this proposed model reaps the benefit of both. Finally, we test this proposed model in difficult scenarios and low-quality license plate images caused by weather, and results show this proposed license plate detection and recognition model can achieve a higher mean average precision, better comprehensive performance, and excellent robustness.

Tracking control is one of the important working conditions of unmanned driving and can help vehicles keep distances and run in an orderly manner to improve traffic utilization, ensure clear roads and avoid rear-end collisions. This paper constructs the motion trend of obstacles within the predictive step length of model predictive control (MPC), designs the danger level indicator between vehicles and obstacles, and formulates dynamic planning for the original reference path based on MPC. This paper proposes an integrated collision avoidance control strategy based on dynamic nonlinear MPC (DNMPC), and predicts the locations of moving obstacles within the predictive step length of DNMPC. The proposed method can perform initial planning for the collision-avoidance path according to the road environment information and based on the activation function. It builds the function for the tendency of obstacles within the predictive step length of MPC, introduces it into the objective function for optimization, designs dynamic path planning control based on the theories of MPC, and performs local optimization to the initial reference path under the obstacles in motion with a mass model. In addition, it defines varying discrete step lengths within the predictive step length and achieves the long-distance prediction and high-precision control of collision avoidance controllers. The experimental results show that the dynamic, nonlinear, and integrated collision avoidance control proposed in this paper can ensure excellent collision avoidance and steady vehicle driving and has very good practical value.

Emerging non-volatile memory (also termed as persistent memory, PM) technologies promise persistence, byte-addressability, and DRAM-like read/write latency. A proliferation of persistent memory systems have been proposed to leverage PM for fast data persistence and expose malloc-like persistent APIs. By eliminating disk I/Os, these systems gain low-latency and high-throughput access performance for persistent data. However, there still exist non-negligible limitations in these systems, such as frequent context switches, inefficient allocation, heavy logging overhead, and lack of wear-leveling techniques. To solve these problems, we develop Lewat, a lightweight, efficient, and wear-aware transactional persistent memory system. Lewat is built in user-layer to avoid kernel/user layer context switches and enables lightweight persistent data access. We decouple the data space into slot zone and page zone. Based on this, we design different allocators in these two zones to achieve efficient allocation performance for both small-sized data and large-sized data. To minimize logging overhead, we propose an efficient adaptive logging framework. The main idea is to utilize different logging techniques for different workloads. We also propose a suite of system-coupled wear-leveling techniques that contain wear-aware allocation, wear-aware update, and write reduction. We evaluate Lewat on a real non-volatile memory platform and the experimental results show that compared with state-of-the-art persistent memory systems, Lewat has much lower latency and higher throughput.

Monitoring various types of disasters involves diversified requirements, such as the spectral band, resolution, and timeliness. However, at present, different types of observation platforms are separately operated. This isolated resource organization model is insufficient to meet the requirements of various Earth-observation tasks, especially when disasters occur. As a result, it is necessary to construct an Earth-observation network that contains space-air-ground observation resources and makes unified task planning for the included heterogeneous resources, such that the efficiency of the entire observation system is maximized. In this article, an architecture with two planning phases is proposed for the coordinated planning of heterogeneous Earth-observation resources, in which area targets and four types of space-air-ground observation resources [i.e., satellites, unmanned aerial vehicles (UAVs), airships, and ground monitoring vehicles] are considered. The two-phase approach in this architecture includes an area target decomposition phase and a task allocation phase. In the first phase, an area target hierarchical decomposition (ATHD) method is proposed to decompose the area targets into subtasks. In the second phase, a task conflict heuristic allocation (TCHA) method is proposed to allocate the decomposed subtasks to subplanning centers. Extensive experiments on simulated and realistic scenarios are conducted to verify the effectiveness of the proposed ATHD and TCHA methods. The computational results show that the ATHD method substantially improves the efficiency of the coordinated task planning process. Moreover, compared with traditional task allocation methods, the TCHA method could produce high-quality observation plans for the Earth-observation network, as it brings complementary benefits via the comprehensive usage of heterogeneous space-air-ground resources.

New and emerging types of Persistent Memory (PM) technologies boost the opportunity to improve the performance of storage systems. PM can unify the main memory and secondary storage by incorporating it into legacy computer systems through the memory bus. In recent years, innovative results have been presented that exploit the byte-addressability, low latency, and non-volatility of PM; these have included local PM file systems and PM systems. However, the high overhead of ensuring data consistency has limited the performance of these systems. In this article, we propose LosPem, a novel log-structured framework for persistent memory to address the performance challenge. LosPem utilizes two techniques to accomplish this. Firstly, LosPem deploys efficient hash-indexed linked lists to maintain the log contents to reduce the significant overhead of log content retrieval. Secondly, LosPem improves the transaction throughput by decoupling a transaction into two asynchronous steps and creating a write buffer on Dynamic Random Access Memory (DRAM) write buffer for processing the frequent data writes. The experimental results show that LosPem outperforms Non-volatile Memory Library (NVML), Mnemosyne and Log-structured Non-volatile Main Memory (LSNVMM) by 27%, 1.2x, and 1.0x on a read-intensive workload. On a write-intensive workload, LosPem outperforms NVML, Mnemosyne, and LSNVMM by 1.8x, 1.2x, and 34%, respectively.

Next-generation Storage Class Memory (SCM) offers low-latency, high-density, byte-addressable access and persistency. The potent combination of these attractive characteristics makes it possible for SCM to unify the main memory and storage to reduce the storage hierarchy. Aiming for this, several persistent memory systems were designed. However, the heavy metadata and transaction cost degrade the system performance. Moreover, neither of them pays attention to wear-leveling strategy. In this article, we present a lightweight persistent memory system, LiwePMS, which allows a fast access to persistent data stored in SCM with wear-aware memory management. LiwePMS makes performance improvement by simplifying the metadata management and the consistency method. LiwePMS abstracts SCM as heap space with container-based dynamic address mapping. Also, LiwePMS implements efficient wear-aware dynamic memory allocator and lightweight transaction mechanism for data consistency in user-space library. The experiments showed that LiwePMS persists key-value records 1.5× faster than Redis RDB mechanism. LiwePMS improves the performance of persistent region operation by more than 45%, 63%, and 1.1× comparing with HEAPO, Mnemosyne, and NVML, respectively. Also, the wear-leveling policy of memory allocator outperforms that of NVMalloc from 35% to 30%, and the transaction method promotes the transaction performance to 1.8× compared to NVML.

Modern Non-Volatile Memory (NVM) promises persistence, byte-addressability and DRAM-like read and write latency, which offers great opportunities for big data storage architecture. These excellent properties indicate that NVM has the potential to be incorporated with key-value stores to achieve high performance and durability simultaneously.

Non-volatile memory(NVM) promises a DRAM replacement in computer systems due to its attractive characteristics. However, the low endurance problem limits its practical applications. In this paper, we propose Wamalloc, an efficient NVM memory allocator to extend the lifetime of NVM in the software level. An elaborate hybrid wear-leveling policy is proposed in this paper to achieve wear-leveling without hardware overhead. The evaluations show that the wear-leveling policy of Wamalloc outperforms that of NVMalloc from 3% to 30%, and the total memory consumption of Wamalloc outperforms that of NVMalloc about 60% and 10% under uniform and random workloads. In addition, the allocation performance of Wamalloc is better than the standard glibc malloc and NVMalloc by 98% and 97% under uniform workloads, 83% and 86% under random workloads.