Traffic flow prediction is one of the core technology of Intelligent Transportation System. Its fundamental challenge is to effectively model the complex spatial-temporal dependencies. Although extensive research has been conducted in this field, the limitations of current methods restrict their effectiveness in accurate predictions. For temporal dependence, existing methods based on recurrent neural networks only focus on local dependencies and ignore global dependencies. For spatial dependencies, existing methods use predefined or adaptive adjacency matrices that cannot accurately reflect the relationships between real traffic flow. To overcome these limitations, we propose a dynamic spatial-temporal fusion graph convolution network (DSTFGCN). In the temporal aspect, we introduce gated dilated causal convolution to capture the local dependencies and node-independent temporal graph convolution to capture the global dependencies specific to each node. In the spatial aspect, we propose a dynamic graph convolution block. It can construct dynamic graphs based on the characteristics of the input data and aggregate both local and global spatial dependencies. Experiments on six real-world datasets have shown that DSTFGCN outperforms current mainstream methods. The codes are available at https://github.com/SYLan2019/DSTFGCN.

In Online Social Networks (OSNs), privacy issue is a growing concern as more and more users are sharing their candid personal information and friendships online. One simple yet effective attack aims at private user data is to use socialbots to befriend the users and crawl data from users who accept the attackers' friend requests. With the attackers involving, individual users' preference and habit analysis is available, hence it is easier for the attackers to trick the users and befriend them. To better protect private information, some cautious, high-profile users may refer to their friends' decisions when receiving a friend request. The aim for this paper is to analyze the vulnerability of OSN users under this attack, in a more realistic setting that the high profile users having a different friend request acceptance model. Specifically, despite the existing probabilistic acceptance models, we introduce a deterministic linear threshold acceptance model for the cautious users such that they will only accept friend requests from users sharing at least a certain number of mutual friends with them. The model makes the cautious users harder to befriend with and complicates the attack. Although the new problem with multiple acceptance models is non-submodular and has no performance guarantee in general, we introduce the concept of adaptive submodular ratio and establish an approximation ratio under certain conditions. In addition, our results are also verified by extensive experiments in real-world OSN data sets.

Device-to-device (D2D) communication has recently gained much attention for its potential to boost the capacity of cellular systems. D2D enables direct communication between devices while bypassing a base station (BS), hence decreasing the load of BSs and increasing the network throughput via spatial reuse of radio resources. However, the cellular system is highly dynamic, an optimal allocation plan of radio resource to D2D links at one time point can easily become suboptimal when devices move. Thus, to maximize spatial reuse in cellular systems, it is crucial to update the resource allocation adaptively to reflect the current system status. In this paper, we develop the first adaptive solution framework to the dynamic resource problem for maximizing spatial reuse. At the core of the framework, we present the two algorithms for the adaptive set multicover problem with approximation ratio f and log n, respectively, where f is the frequency of the most frequent element and n is the total number of elements. The experimental results not only show that our solutions have a significant improvement in running time, compared with optimal or approximated offline methods, but also demonstrate their good performance through the resource usage, network throughput, and other metrics.

The vulnerability of interdependent networks has recently drawn much attention, especially in the key infrastructure networks such as power and communication networks. However, the existing works mainly considered a single cascade model across the networks and there is a need for more accurate models and analysis. In this paper, we focus on the interdependent power/communication networks to accurately analyze their vulnerability by considering heterogeneous cascade models. Accurately analyzing interdependent networks is challenging as the cascades are heterogeneous yet interdependent. Also, including multiple timescales into the context can further increase the complexity. To better depict the vulnerability of interdependent networks, we first propose a method to learn a threshold model from historical data to characterize the cascades in the power network and alleviate the need of calculating complicated power network dynamics. Next, we introduce message passing equations to generalize the threshold model in the power network and the percolation model in the communication network, based on which we derive efficient solution for finding the most critical nodes in the interdependent networks. Removing the most critical nodes can cause the largest cascade and thus characterizes the vulnerability. We evaluate the performance of the proposed methods in various datasets and discuss how network parameters, such as the timescales, can impact the vulnerability.

Online Social Networks (OSNs) are effective platforms for viral marketing. Due to their importance, viral marketing related problems in OSNs have been extensively studied in the past decade. However, none of the existing works can cope with the situation that the propagation rate dynamically increases for popular topics, as they all assume known propagation rates. In this paper, to better describe realistic information propagation in OSNs, we propose a novel model, Dynamic Influence Propagation (DIP), that allows propagation rate to change during the diffusion. We then define a new research problem: Threshold Activation Problem under DIP (TAP-DIP) to study the impact of DIP. TAP-DIP adds extra complexity on the already #P-hard TAP problem. Despite it hardness, we are able to approximate TAP-DIP with O(log|V|) ratio. Sitting in the core of our algorithm are the Lipschitz optimization technique and a novel solution to the general version of TAP, the Multi-TAP problem. Using various real OSN datasets, we experimentally demonstrate the impact of DIP and that our solution not only generates high-quality seed sets when being aware of the rate increase, but also is scalable.

We consider the Threshold Activation Problem (TAP): given social network G and positive threshold T, find a minimum-size seed set A that can trigger expected activation of at least T. We introduce the first scalable, parallelizable algorithm with performance guarantee for TAP suitable for datasets with millions of nodes and edges; we exploit the bicriteria nature of solutions to TAP to allow the user to control the running time versus accuracy of our algorithm through a parameter α ϵ (0, 1): given η > 0, with probability 1 - η our algorithm returns a solution A with expected activation greater than T - 2αT, and the size of the solution A is within factor 1-h 4α/T + log(T) of the optimal size. The algorithm runs in time O (α-2log (n/η) (n + m)|A|), where n, m, refer to the number of nodes, edges in the network. The performance guarantee holds for the general triggering model of internal influence and also incorporates external influence, provided a certain condition is met on the cost-effectivity of seed selection.

Device-to-device (D2D) communications over the licensed wireless spectrum has been recently proposed as a promising technology to meet the capacity crunch of next generation cellular networks. However, due to the high mobility of cellular devices, establishing and ensuring the success of D2D transmission become a major challenge. To this end, in this paper, a novel framework is proposed to enable devices to form multi-hop D2D connections in an effort to maintain sustainable communication in the presence of device mobility. To solve the problem posed by device mobility, in contrast to existing works, which mostly focus on physical domain information, a durable community-based approach is introduced taking social encounters into context. It is shown that the proposed scheme can derive an optimal solution for time sensitive content transmission while also minimizing the cost that the base station pays in order to incentivize users to participate in D2D. Simulation results show that the proposed social community aware approach yields significant performance gain, in terms of the amount of traffic offloaded from the cellular network to the D2D tier, compared with the classical social-unaware methods.