Cheng Wang

Time-Sensitive Networking (TSN) is an advance Ethernet paradigm designed to provide low delay, low jitter, and deterministic transmission time. The Cycling Queuing and Forwarding (CQF) mechanism is introduced in TSN as a scheduler to achieve precise communication. Multi-CQF, as an extension of CQF, supports the transmission of various traffic types by assigning different cycle lengths to each queue group. In its original form, Multi-CQF-based scheduling algorithms do not account for flow sorting, leading to increased transmission delays and reduced network efficiency as a network dynamically changes. To enhance the performance of Multi-CQF, this paper initially utilizes queuing theory to analyze and manage traffic, providing foundation solutions. Subsequently, Mixed Integer Programming (MIP) and the Variable Neighborhood Search Genetic Algorithm (VNS-GA) are employed to optimize transmission delay in small- and large-traffic TSN networks, respectively. MIP quickly seeks out the optimal scheduling solution for small-traffic TSN networks using branch-and-bound and linear programming techniques, while the VNS-GA improves efficiency and performance for large-traffic ones by continuously adjusting the search neighborhood strategy. Comparing with other existing schemes, computer simulation reveals that MIP reduces delay by approximately 13% on average in small-traffic TSN networks, while the VNS-GA achieves an average delay reduction of 7% in large-traffic ones.

Flow admission control (FAC) aims to efficiently manage the service requests while maximizing the network utilization. With multiple connection requests, access delay or even service interruption may occur. This paper proposes a novel FAC approach to reduce the contention between the end nodes and ensure high utilization of the networking resources for software defined IIoT. First, incoming flows are classified into different priorities using back propagation neural network based on selected features representing the current network status. Second, with the designed flow admission policies, bandwidth and buffer size are estimated with stochastic network calculus model. Finally, the thresholds of the proposed FAC scheme are dynamically decided based on the above two parameters. Various flows are admitted or rejected via the proposed FAC to maintain real time processing. Unlike traditional FAC schemes rely on static priority systems, the proposed scheme leverages machine learning technique for dynamic flow prioritization and the stochastic network calculus model for precise estimation. Computer simulation reveals that the proposed scheme accurately classifies the flows, and substantially decreases the transmission delay and improves the network utilization compared to the existing FAC schemes. This highlights the superiority of the proposed scheme meeting the demands of software defined IIoT.

With the advancement of Industry 4.0, Time-Sensitive Networking (TSN) has become essential for ensuring prompt and reliable data transmission. As an augmentation of Ethernet, TSN aims to supply services capable of low latency, minimal jitter, and low packet loss for urgent data in decentralized, user-oriented networks. Efficient detection techniques are integral to TSN for swiftly determining the practicability of network configurations, as existing schedulability analysis proves insufficient. This paper delves into the potential of backpropagation neural networks (BPNN) in schedulability analysis efficiency. We optimize BPNN using spearman correlation feature selection combined with a voting ensemble method and Particle Swarm Optimization, forming two models: Spearman-Vote-BPNN and Spearman-PSO-BPNN. Testing on 5,000 network configurations in computer simulations, both models demonstrated high generalization accuracy, around 97.4%. Spearman-Vote-BPNN achieved the fastest training speed at 0.63 seconds and an accuracy of 98.2%. Meanwhile, Spearman-PSO-BPNN showed the highest accuracy (98.5%) with the quickest detection speed (5.6ms). The outcomes of this research significantly advance the efficacy and precision of TSN network configuration detection and establish a formidable groundwork for future scholarly pursuits in this area.

As the extension of cloud computing, edge computing has witnessed remarkable progress in the last few years. In edge computing, task offloading from edge device to edge server is becoming a promising technique that substantially enhances the quality of service for edge devices. Nevertheless, appropriate edge server selection for task offloading remains a challenge if multiple edge servers are selectable nearby. In this paper, we propose a parameter estimation-aided edge server selection scheme for task offloading (PEESS) with multiple potential edge servers. The proposed scheme jointly leverages Baum-Welch algorithm and forward algorithm of hidden Markov model. That is, PEESS leverages Baum-Welch algorithm to estimate the parameters of potential edge servers based on their time-sequential incoming task records. Subsequently, the forward algorithm is utilized for the probability calculation of the observations based on the estimated parameters. Eventually, the probability calculation results are sorted in descending order, and the edge device selects the target edge server based on the sequence of sending requests to the available edge servers from optimal to suboptimal. Extensive simulation results demonstrate that PEESS outperforms other conventional and state-of-the-art edge server selection schemes in terms of time consumption and energy consumption.

In recent years, much research has been conducted on time series based human activity recognition (HAR) using wearable sensors. Most existing work for HAR is based on the manual labeling. However, the complete time serial signals not only contain different types of activities, but also include many transition and atypical ones. Thus, effectively filtering out these activities has become a significant problem. In this paper, a novel machine learning based segmentation scheme with a multi-probability threshold is proposed for HAR. Threshold segmentation (TS) and slope-area (SA) approaches are employed according to the characteristics of small fluctuation of static activity signals and typical peaks and troughs of periodic-like ones. In addition, a multi-label weighted probability (MLWP) model is proposed to estimate the probability of each activity. The HAR error can be significantly decreased, as the proposed model can solve the problem that the fixed window usually contains multiple kinds of activities, while the unknown activities can be accurately rejected to reduce their impacts. Compared with other existing schemes, computer simulation reveals that the proposed model maintains high performance using the UCI and PAMAP2 datasets. The average HAR accuracies are able to reach 97.71% and 95.93%, respectively.