Fast charging stations are critical infrastructures to enable high penetration of plug-in electric vehicles (PEVs) into future distribution networks. They need to be carefully planned to ensure meeting charging demand as well as economic benefits. Accurate estimation of PEV charging demand is the prerequisite of such planning, but a non-trivial task. This paper addresses the sizing (number of chargers and waiting spaces) problem of fast charging stations and presents an optimal planning solution based on an explicit temporal-SoC characterization of PEV fast charging demand. The characteristics of PEV charging demand are derived through the vehicle travel behavior analysis using available statistics. The PEV dynamics in charging stations is modelled with a Markov chain and queuing theory. As a result, the optimal number of chargers and waiting spaces in fast charging stations can be jointly determined so as to maximize the expected operator profits, considering profit of charging service, penalty of waiting and rejection, as well as maintenance cost of idle facilities. The proposed solution is validated through a case study with mathematical justifications and numerical results from simulation.

Vehicular Ad-Hoc Networks (VANETs), as the crucial support of Intelligent Transportation Systems (ITS), have received great attention in recent years. With the rapid development of VANETs, various services have generated a great deal of data that can be used for transportation planning and safe driving. Especially, with the advent of Coronavirus Disease 2019 (COVID-19), the transportation system has been impacted, thus novel modes of transportation planning and intelligent applications are necessary. Digital twins can provide powerful support for artificial intelligence applications in Transportation Big Data (TBD). The features of VANETs are varying, which arises the main challenge of digital twins applying in TBD. Network traffic prediction, as part of digital twins, is useful for network management and security in VANETs, such as network planning and anomaly detection. This paper proposes a network traffic prediction algorithm aiming at time-varying traffic flows with a large number of fluctuations. This algorithm combines Deep Q-Learning (DQN) and Generative Adversarial Networks (GAN) for network traffic feature extraction. DQN is leveraged to carry out network traffic prediction, in which GAN is involved to represent Q-network. Meanwhile, the generative network can increase the number of samples to improve the prediction error. We evaluate the performance of our method by implementing it on three real network traffic data sets. Finally, we compare the two state-of-the-art competing methods with our method.

Artificial intelligence (AI) is a very powerful technology and can be a potential disrupter and essential enabler. As AI expands into almost every aspect of our lives, people raise serious concerns about AI misbehaving and misuse. To address this concern, international organizations have put forward ethics guidelines for constructing trustworthy AI (TAI), including privacy, transparency, fairness, robustness, accountability, and so on. However, because of the black-box characteristics and complex models of AI systems, it is challenging to translate these guiding principles and aspirations into AI systems. Blockchain, an important decentralized technology, can provide the capabilities of transparency, traceability, immutability, and secure sharing and hence can be used to make AI trustworthy. In this paper, we survey studies on blockchain-based TAI (BTAI) from a software development lifecycle view. We classify the lifecycle of BTAI into four stages: Planning, data collection, model development, and system deployment/use. Particularly, we investigate and summarize the trustworthy issues that blockchain can achieve in the latter three stages, including (1) data transparency, privacy, and accountability; (2) model transparency, privacy, robustness, and fairness; and (3) robustness, privacy, transparency, and fairness of system deployment/use. Finally, we present essential open research issues and future work on developing BTAI systems.

In multi-access edge computing systems with energy harvesting (MEC-EH), the mobile devices are empowered with unstable energy harvested from renewable energy sources. To prolong the life of mobile devices, as many computation-intensive tasks as possible should be offloaded to the MEC server. However, when the system states of mobile device and MEC server are unstable, e.g. poor communication channel conditions, a great number of tasks will be executed locally, leading to a long execution time. Even worse, some tasks may be dropped due to low energy levels. To address this problem, in this paper, we propose a multi-relay assisted computation offloading framework for MEC-EH systems. In this framework, a computation task can be executed by offloading to the MEC server with the help of multiple relay nodes, such as the neighboring nodes. We introduce execution cost as a performance metric to incorporate both the task execution time and task failure. We then develop a low-complexity online algorithm, namely MRACO algorithm, to minimize the average execution cost. MRACO algorithm can select the optimal execution strategy for each task from (1) executing the task locally, (2) offloading it to the MEC server directly, (3) offloading it to the MEC server with the help of the most suitable neighboring nodes, and (4) simply dropping it. Moreover, we also develop an algorithm for selecting the suitable neighboring devices to act as relays and determining the optimal task splitting ratio between them. Finally, performance evaluation shows that the proposed MRACO algorithm greatly outperforms the benchmarks in terms of both average execution time and task drop rate.