Traffic prediction is vital to traffic planning, control, and optimization, which is necessary for intelligent traffic management. Existing methods mostly capture spatiotemporal correlations on a fine-grained traffic graph, which cannot make full use of cluster information in coarse-grained traffic graph. However, the flow variation of clusters in the coarse-grained traffic graph is more stable compared with nodes in the fine-grained traffic graph. And the flow variation of a fine-grained node is generally consistent with the trend of the cluster to which the node belongs. Thus information in the coarse-grained traffic graph can guide feature learning in the fine-grained traffic graph. To this end, we propose a Spatiotemporal Multiscale Graph Convolutional Network (SMGCN) that explores spatiotemporal correlations on a multiscale graph. Specifically, given a fine-grained traffic graph, we first generate a coarse-grained traffic graph by graph clustering, and extract spatiotemporal correlations on both fine-grained and coarse-grained traffic graphs. Then we propose a cross-scale fusion (CF) to implement information diffusion between the fine-grained and coarse-grained traffic graphs. Moreover, we employ an adaptive dynamic graph convolution network to mine both static and dynamic spatial features. We evaluate SMGCN on real-world datasets and obtain a $1.18\% -3.32\%$ improvement over state-of-the-arts.

The Internet of Things is one of the new emerging application domains that require delay tolerant network (DTN) support, where an end-to-end path between the source and the destination may not always exist. Due to the intermittent connectivity of DTN, the design of an efficient routing algorithm is the main challenge. In this paper, we first define a metric called message handling capacity to determine the ability of a node to forward messages. Then, we introduce a concept called connection strength to reflect the connection time between nodes and then integrate the concept into delivery predictability used by Prophet to determine the chance of a node completely delivering a message to the destination. Subsequently, we present a metric called quality of node (QoN), which is calculated by combining the relative weights of the message handling capacity and the improved delivery predictability. Finally, we present an adaptive spray and wait routing algorithm based on QoN (QoN-ASW). The QoN-ASW adaptively allocates the number of message copies between the encountered nodes according to the proportion of quality of node in the spray phase, which avoids the blindness of replica distribution. In addition, a forwarding scheme is implemented in the wait phase, which takes full advantage of encounter opportunities. In the simulation, we demonstrate the efficiency of integrating the connection strength into delivery predictability and compare the QoN-ASW with four existing DTN routing algorithms from four aspects. The simulation results show that the QoN-ASW can significantly improve the delivery rate and reduce the average delay while achieving a relatively low overhead.

Vehicular Cloud and autonomous vehicles require a scalable and reliable mobile communication network. LTE and Dedicated Short Range Communication (DSRC) have been trying to fit for such role, yet neither can satisfy all requirement due to inherent architectural limitations. Fortunately the fifth generation mobile network, 5G and Mobile Edge Cloud/Computing (MEC) is around the corner, targeting ultra low packet delay, high reliability, and Gigabit level wireless bandwidth. This paper introduces a unique vehicular MEC architecture where instead of simply off-loading application service to the edge servers on MEC, vehicular communication packets are routed through the MEC network. We discuss in detail how it accommodates vehicle to vehicle (V2V) and vehicle to infrastructure (V2I) communication with high scalability and guaranteed low packet delay. We also provide an in depth analysis of the pros and cons of our MEC vehicle network design, and address the mobility management issue on edge cloud. Applying distributed mobility management (DMM) operations we are able to make edge cloud IP handoff seamless and transparent. Proof of concept simulations are conducted using NS3.

With the rapid development of smart cities, the collection of vehicle trajectory data through sensors has increased significantly. While many studies have utilized calibrated physical car-following models (CFM) and machine learning techniques for trajectory prediction, these approaches often falter in complex, dynamic traffic scenarios. Addressing this gap, this paper introduces PS-TrajGAIL, a generative adversarial imitation learning framework tailored for urban vehicle trajectory generation. Contrary to conventional discriminative models, PS-TrajGAIL employs a generative model to capture the inherent distribution of urban vehicle trajectories. This framework models the tasks of trajectory generation as a partially observable Markov decision process based on imitation learning. PS-TrajGAIL's architecture features a generator, which simulates vehicle behavior to produce synthetic trajectories, and a discriminator that distinguishes between authentic and generated trajectories. In addition, the driving policy within the generator is fine-tuned using the Trust Region Policy Optimization (TRPO) algorithm, ensuring safety in vehicle driving. Experimental evaluations on both synthetic and real-world datasets highlight that PS-TrajGAIL notably surpasses existing baselines and state-of-the-art approaches in trajectory generation

SUMMARY: The use of multiple channels in 802.11 wireless local area networks can improve network performance. Many efforts have been done to better exploit multiple non‐overlapped channels. However, the number of orthogonal channels in the Institute of Electrical and Electronics Engineers 802.11 standards is very much limited. Recent studies indicate that we can improve the full‐range channel utilization and the network throughput by properly utilizing the partially overlapping channels. However, little work was focused on channel assignment for partially overlapping channels. In this paper, we investigate the problem of partially overlapping channel assignment to improve the performance of 802.11 wireless networks based on the Signal to Interference–Noise Ratio interference model. Using the Signal to Interference–Noise Ratio model, we deduce a direct relationship between maximizing system throughput and minimizing total interference when partially overlapping channels are employed. After that, we propose a greedy method to minimize the total interference for throughput maximization. We evaluate our algorithm through extensive simulations and compare its performances with those of the state‐of‐the‐art. Copyright © 2013 John Wiley & Sons, Ltd.

In opportunistic network, the regular connectivity of nodes and their individual conditions significantly influence message transmission. To address this, we propose an improved spray and wait routing algorithm based on stable transmission capacity evaluation of nodes (referred to as STCESW), which is derived from traditional Prophet and spray and wait algorithms but surpasses them. STCESW incorporates network models that consider node connection stability and transmission capacity, aiming to enhance the update and punishment formulas for node delivery probability. Furthermore, it combines the residual message buffer size of nodes with the angle between motion directions of nodes to determine more efficient and extensive message spraying in the network. The simulation results clearly indicate that STCESW significantly outperforms the other three enhanced algorithms in both SPMBM and RWP. It demonstrates its robust performance in terms of delivery probability, showing an increase ranging from 2% to 87% across diverse node buffer sizes or simulation time intervals. Additionally, it compromises slightly on delivery latency while maintaining relatively stable message latency and buffer time.