Wei Cheng

Understanding curbside parking rules is crucial for drivers to quickly find legal curbside parking. Traditional data providers rely on manual methods to collect information about curbside parking signs, either by individually noting down sign details or downloading street-level imagery from digital maps. However, this process is slow and inadequate for building an accurate and up-to-date curbside parking rule database, considering the frequent updates to parking signs. In this article, we propose a novel deep-learning-based Inventory Management System (IMS) that automates to create and manage the curbside parking rule database using videos captured by off-the-shelf dashcams installed on vehicles. To the best of our knowledge, our system is the first of its kind to detect and interpret real-world curbside parking signs from videos and generate parking rules. By utilizing a serverless cloud architecture on AWS, IMS combines secure data retrieval, robust user authentication, and responsive map-based visualization, providing users with up-to-date parking information filtered by location, date, and time. Through real-world evaluations, we demonstrate that our system efficiently constructs an accurate curbside parking rule database enhancing urban mobility through efficient and informed parking management.

As autonomous vehicles become more prevalent in urban settings, the need for accurate perception systems capable of high-precision localization and object detection in complex urban environments becomes increasingly critical. Traditional methods often struggle in underground structures or densely built urban zones where sensor data is obstructed or noisy. To address these challenges, we propose a novel integration of multiple cameras and radars, leveraging their complementary strengths to enhance overall perception accuracy. This approach moves beyond the limitations of single-sensor systems, combining the detailed visual data from cameras with the robust distance and velocity measurements from radars. Our innovative framework, MultiFusionNet, incorporates deep learning, adaptive multimodal fusion, and multiscale attention mechanisms to dynamically adjust to varying environmental conditions and complex sensor data interactions. Additionally, we introduce the scene-aware spatiotemporal network (SA-STNet) to refine radar data and improve its resolution and temporal coherence. By optimizing sensor placement with an uncertainty-aware probabilistic metric, we achieve comprehensive environmental coverage and significantly enhance the reliability of autonomous navigation. By optimizing sensor placement with an uncertainty-aware probabilistic metric, we achieve comprehensive environmental coverage, leading to improved detection accuracy (e.g., 0.88 AP and 0.90 recall) in simulated urban scenarios, paving the way for enhanced autonomous navigation reliability pending real-world validation. Validated through extensive simulations in the CARLA-generated datasets, our framework demonstrates notable improvements in detection accuracy, offering a robust solution for navigating complex and challenging urban landscapes.

Sharing the big data generated by IoT via cloud is slow and expensive. Besides, transmitting and sharing data among IoT devices via cloud may be insecure. To address these issues, a novel efficient and secure data sharing scheme termed EB-SDSS (Edge Blockchain Secure Data Sharing Scheme) is proposed in this paper for edge-enabled IoT applications. EB-SDSS constructs a blockchain on edge servers. It guarantees the confidentiality and unforgeability of data by combining the symmetric encryption scheme with an edge blockchain. To ensure the device authenticity and the reliability of shared data, EB-SDSS adopts a certificateless signature scheme. It also provides efficient large-scale data searches for IoT devices through a locality-sensitive hashing algorithm. EB-SDSS has been proven to be secure against the adaptive chosen message attacks under the random oracle model. The experimental results indicate that EB-SDSS is feasible for IoT inter-device data sharing.

Autonomous vehicles could fundamentally alter the way we travel as they offer an unprecedented level of comfort and safety on the roads, including the ability to navigate and park in an appropriate spot. To enable this, robust turn-by-turn navigation tools and curbside parking planning tools are essential to ensure safe and efficient navigation, as they require the vehicle to understand and adhere to parking regulations and to safely and accurately park in a designated spot. In this article, we present a novel smart curbside parking system that includes a turn-by-turn navigation curbside parking searching system, specifically designed to empower the safe navigation of autonomous vehicles to find curbside parking, as well as a curbside parking planning tool designed for users to search for and plan for curbside parking ahead of time. The proposed system utilizes a combination of GPS, map data, and integrates the vehicle routing algorithms to provide real-time, turn-by-turn navigation instructions to the vehicle. The curbside parking rule data used in the proposed system is extracted from the street videos collected utilizing a dashcam mounted on a vehicle. The performance of the proposed system was evaluated on the real-world data set and found capable to significantly reduce the curbside parking searching time during peak hours.

Most data center services are finished by the cooperation among the connected servers. However, the malicious attackers always try to divide the network into disconnected components to start some attacks, such as the address resolution protocol (ARP) attack, the denial of service (DoS) attack, the botnet attack, and so on. The connectivity is an excellent indicator to measure the reliability and fault-tolerant ability of the network. Whereas, the traditional connectivity and current conditional connectivity cannot well reflect the fault-tolerant performance of the network when attackers are a block or have a certain structure and the components of the remaining network still have a certain structure. Based on this fact, we propose a new measure: the double-structure connectivity, which can accurately reflect the fault-tolerant ability of the network when attackers are structured and each component of the network has a certain structure after removing the attacked servers. Meanwhile, a hypercube is a high-performance interconnection network that can also be used to design some data center networks. Therefore, we study the double-structure fault-tolerance of the hypercube and determine the double-structure connectivity of distinct structures of the hypercube. Furthermore, we propose algorithms to construct structures of attackers directly to measure the fault-tolerant ability of the hypercube under this attack. Our results can be applied not only to interconnection networks but also to some data center networks.

Currently, communications in the Vehicular ad hoc network (VANET) can via both Dedicated Short Range Communication (DSRC) and mobile cellular networks. To make use of existing Long Term Evolution (LTE) network in data transmissions, many methods are proposed to manage VANETs. Grouping the vehicles into clusters and organizing the network by clusters is one of the most universal and most efficacious ways. Since the high mobility of vehicles makes VANETs different from other mobile ad hoc networks (MANETs), the previous cluster-based methods for MANETs may have trouble for VANETs. In this paper, we introduce a center-based clustering algorithm to help self-organized VANETs forming stable clusters and decrease the status change frequency of vehicles on highways. A simulation is conducted to compare our mechanism to some other mechanisms. The results show that our mechanism obtains high stability and low packet loss rate.

Ultraviolet (UV) radiation has a great impact on human health. Nowadays, the public basically gets information about UV radiance through weather forecasts. However, weather forecasts just provide rough and average predication for a certain large region. Since CMOS sensors in mobile phone cameras are very sensitive to UV, mobile phones have potential to be ideal equipment to measure UV radiance. This paper introduced a method that can measure UV radiance by just using mobile phone cameras. In addition, by utilizing fog computing, results can be gathered and amended locally through fog server to provide relatively accurate UV measurement. In this paper, advantages of using mobile phones to do UV measurement were discussed at first. Then, theoretical foundations were meticulously illustrated. Later, a procedure that can be implemented in mobile phones was provided. Furthermore, an Android app called UV Meter was developed based on the procedure. At last, verification was conducted under different weather conditions. Results showed that the procedure is valid and can be easily implemented onto mobile phones for everyday UV measurement.

Radio-frequency identification (RFID) systems on roads (RSR) is a recently developed framework that focuses on enhancing transportation safety. In RSR, large-scale passive RFID tags with road-related information are deployed on road surfaces or roadside units. A vehicle with an onboard RFID reader can acquire road information via reading from these RFID tags. As a result, it is critical to update the tags with the latest road information (particularly emergent alerts) in a timely manner in RSR. In this paper, we design a novel cluster-based information diffusion scheme that aims to quickly and accurately update the contents of the RFID tags. To the best of our knowledge, our work is the first to focus on the problem of updating RFID tags in RSR. Our proposed scheme organizes RFID tags into clusters. The emergent information and update status of these clustered tags can be read by vehicles passing by. These vehicles can exchange such information with each other via vehicle-to-vehicle (or vehicle-to-infrastructure) (V2V or V2I) communications. After synthesizing the received information, these vehicles can update the nearby tags and spread the information accordingly. Our extensive simulation results demonstrate that our proposed information diffusion scheme can effectively update the tags within a certain time constraint under various scenarios.