Ying Li

Accurate causal discovery in telecommunication alarm event sequences is crucial for reliable root cause analysis, but presents significant challenges due to complex topological dependencies and inherent alarm redundancy. To address these challenges, we propose the topological-temporal convolution transformer Hawkes process (TTCTH), a novel framework for learning hidden causal relationships in alarm event sequences. To start with, a topological-temporal graph convolution is designed to initialize a causal graph for mainly exploring local dependencies among event types in the topological and temporal domains. We next design three separable transformer variants of topological transformer (TopoT), causal transformer and temporal transformer, which are sequentially stacked to globally capture causal dependencies among event types. In particular, the TopoT fuses prior topological information into alarm sequence representation. The causal transformer models causal dependencies with self-attention mechanism to capture spatially causal relations among event types. The temporal transformer is developed to model long-range causal dependencies across multiple events. We then leverage the high-dimensional hidden state representation, encoded by the stacked transformers, to construct feature evolution with the constraint-based Hawkes process. In order to observe the performance of TTCTH, we conduct extensive experiments on real-world alarm datasets. It demonstrates TTCTH’s excellence in comparison with ten baselines.

Computation-intensive vehicle tasks sharply increase with the rapid growth of intelligent vehicles. The technology of Mobile Edge Computing (MEC) has the possibility of assisting vehicles with computation offloading. To solve the problem of computation resource management and guarantee the security of resource transaction, we jointly combine the MEC network and the blockchain networks to build a blockchain based MEC offloading model. The non-cooperative interactions between MEC server and vehicles formulate a two-stage Stackelberg game in an aim to maximize their benefits and information security. We theoretically demonstrate the unique existence of Nash equilibrium, which enables participants to decide their optimal strategies. Finally, the performance of the proposed model is analyzed by conducting simulation experiments. Our proposed model optimizes resource allocation and also improves the security of the whole network.

The prediction of traffic conditions plays a significant role in the intelligent transportation system. Since the traffic flow is able to directly reflect traffic situation, it is of great value to predict the traffic flow in reality. The traffic flow data has complex spatiotemporal characteristics. Meanwhile, a high node correlation between nodes in the traffic network. Therefore, the key point of accurate traffic flow prediction is to analyze the temporal and spatial correlation in the data, as well as integrate the node correlation. We are motivated to propose a spatial-temporal node correlation network (STNCN) for improving traffic flow prediction accuracy. First, a node correlation matrix is constructed to describe the pairwise correlation between nodes. We then propose a recurrent graph network to analyze node correlation and spatial correlation of traffic flow data. Lastly, a temporal convolutional network is applied to extract temporal features. We combine the outputs of the temporal convolutional networks of different periods into the fully connected layer to obtain final prediction results. The prediction results on real datasets demonstrate the effectiveness of the proposed method.

Activity prediction aims to predict what activities will occur in the future. In smart homes, to facilitate the daily living of the residents, automated or assistive services are provided. To provide these services, the ability of activity prediction is necessary. When we make a prediction, most of the existing works focus on predicting information about the next activity. However, in a smart home environment, compared with just predicting information about the next activity, another type of activity prediction problem has more practical value: predicting what activities will occur in the coming time period of a certain length. The necessity of this type of prediction is due to the purpose of the smart homes and the character of the activities. Many activities in smart homes need preparation time before being performed. Through this type of prediction, activities could be predicted sufficient time before being performed, and there will be adequate time for the smart home system to prepare corresponding automated or assistive services. As more than one activity could occur within the time period in which the prediction is made, this problem is a multilabel classification problem. In this article, we first give a formulation of the problem. Then, we propose a deep learning model to address it. The proposed model consists of the convolutional part, the long short-term memory layer, and the multilabel output layer. Experiments on real-world datasets show the effectiveness of our model.

Blockchain technology is a promising resource management architecture due to its ability of building trust in a decentralized transaction. Block mining participants, i.e. miners, are incentivized with reward for successfully mining blocks. Unfortunately, solving the proof-of-work puzzle consumes substantial computing powers during the mining period, which greatly challenges miners. Mobile devices also fail to participate in mining because of limited resource. To solve these issues, we are motivated to propose a mining framework of alleviating miner’s computation-intensive mining burdens, as well as enabling mobile devices’ participation. Depending on the proposed model, miners are capable of offloading their computation-intensive tasks to the edge cloud and mobile devices. The interactions among them formulate a muti-leader multi-follower Stackelberg game. We achieve the Subgame Perfect Equilibrium (SPE) in the game, which guarantees three types of participants to realize profit maximization. Simulation results demonstrate the effectiveness of the proposed model.