Hypergraph learning is a technique conducting learning on a hypergraph structure. In recent years, hypergraph learning has attracted increasing attention due to its flexibility and capability in modeling complex data correlation. In this paper, we first systematically review existing literature regarding hypergraph generation, including distance-based, representation-based, attribute-based, and network-based approaches. Then we introduce the existing learning methods on a hypergraph, including transductive hypergraph learning, inductive hypergraph learning, hypergraph structure updating, and multi-modal hypergraph learning. After that, we present a tensor-based dynamic hypergraph representation and learning framework that can effectively describe high-order correlation in a hypergraph. To study the effectiveness and efficiency of hypergraph generation and learning methods, we conduct comprehensive evaluations on several typical applications, including object and action recognition, Microblog sentiment prediction, and clustering. Besides, we contribute a hypergraph learning development toolkit called THU-HyperG.

The digital transformation of industry requires industrial control networks provide high flexibility and determinacy. Time-Sensitive Software-Defined Networking (TSSDN) that combines Time-Sensitive Networking (TSN) and Software-Defined Networking (SDN) is a new network paradigm which provides both real-time transmission feature and network flexibility. During network updates, the transmission consistency needs to be maintained. However, previous mechanisms mostly target on the proper schedule transition, which cannot guarantee no frame loss and also introduces extra update overhead. The paper proposes a novel flow schedule generation model which guarantees no frame loss during network updates even with the basic two-phase update mechanism and introduces no extra update overhead. Two algorithms are designed for the model to adapt to different application scenarios: the offline algorithm poses better schedulability, whereas the online one consumes less time with slightly decreased schedulability. The experiments on two real-world industrial networks demonstrate our mechanism achieves zero frame loss without extra update overhead compared to existing methods, and the online algorithm saves 40% execution time with at most 10% schedulability decrease when the bandwidth utilization is less than 50%.

With the development of hardware virtualization and cloud computing, modern industry has a tendency to upgrade from the traditional industrial networks to virtual machine (VM) based networks. To provide firm latency guarantees for control messages in these networks, time-sensitive network (TSN) is a promising technology due to its determinacy for real-time applications. However, TSN faces the challenge of providing a rapid response to dynamic transmission requirement changes incurred by VM migrations. In this paper, We proposed an online scheduling approach to deal with dynamic VM migrations in multicast TSN. In this approach, we devise a novel online scheduling framework (MDTC-HB) containing an offline scheduling phase and an online rescheduling phase. While the offline phase introduces a minimal distance tree (MDT) to increase reusable scheduling results, the online phase proposes a heuristic scheduling approach to reuse the results of the offline phase as much as possible to accelerate the rescheduling process. Experiments show that our framework can provide a rapid response to dynamic VM migrations compared with the existing approaches where the amount of control data does not exceed 50% of the bandwidth.

This paper studies the important sensory data collection problem in the sensor-augmented RFID systems, which is to quickly and accurately collect sensory data from a predefined set of target tags with the coexistence of unexpected tags. The existing RFID data collection schemes suffer from either low time-efficiency due to tag-collisions or serious data corruption issue due to interference of unexpected tags. To overcome these limitations, we propose the hierarchical-hashing data collection (HDC) protocol, which can not only significantly improve the utilization of RFID wireless communication channel by establishing bijective mapping between k target tags and the first k slots in time frame, but also effectively filter out the serious interference of unexpected tags. Although HDC has attractive advantages, the theoretical analysis reveals that the computation cost involved in it is as huge as O(k2k), where k is normally large in practice. By making some modifications to the basic HDC protocol, we propose the multi-framed hierarchical-hashing data collection (MHDC) protocol to effectively reduce the involved computation complexity. Unlike HDC that only issues a single time frame, MHDC uses multiple time frames to collaboratively collect sensory data from the k target tags. It can be understood as that a big computation task is disintegrated into multiple small pieces and then shared by multiple time frames. As a result, the computation cost involved in MHDC is reduced to O(k2n), where nłl k is the expected number of target tags that each time frame handles. Theoretical analysis is given to jointly consider the communication cost and computation cost thereby maximizing the overall time-efficiency of MHDC. Extensive simulation results reveal that the proposed MHDC protocol can correctly collect all sensory data and is always about more than 2x faster than the state-of-the-art RFID sensory data collection protocols.

The execution time of conventional incremental off-line schedule approaches for time-triggered networks increases rapidly when networks become larger. When the traffic in a network changes, they need to reschedule all influenced flows once again. Traffic changes at the cluster level involve many data flows. An incremental scheduler cannot react quickly to such changes. We propose an algorithm based on mixed integer linear programming and counterexample guided methodology. Our algorithm can generate adaptive schedule for cluster-level changes of the system. The adaptive schedule can react quickly to the changes during runtime. Our algorithm enhances the incremental schedulers. It allows schedulers to react to changing at both the flow level and the cluster level. Experiments show that our approach is effective. In the scenarios of coupling train consists, our algorithm can generate the schedule table of the train network within a few seconds.

Complex and customized manufacturing requires a high level of collaboration between production and logistics in a flexible production system. With the widespread use of Internet of Things technology in manufacturing, a great amount of real-time and multi-source manufacturing data and logistics data is created, that can be used to perform production-logistics collaboration. To solve the aforementioned problems, this paper proposes a timed colored Petri net simulation-based self-adaptive collaboration method for Internet of Things-enabled production-logistics systems. The method combines the schedule of token sequences in the timed colored Petri net with real-time status of key production and logistics equipment. The key equipment is made ‘smart’ to actively publish or request logistics tasks. An integrated framework based on a cloud service platform is introduced to provide the basis for self-adaptive collaboration of production-logistics systems. A simulation experiment is conducted by using colored Petri nets (CPN) Tools to validate the performance and applicability of the proposed method. Computational experiments demonstrate that the proposed method outperforms the event-driven method in terms of reductions of waiting time, makespan, and electricity consumption. This proposed method is also applicable to other manufacturing systems to implement production-logistics collaboration.