1 January 2021 in Wireless Communications and Mobile Computing
The demand for improving productivity in manufacturing systems makes the industrial Internet of things (IIoT) an important research area spawned by the Internet of things (IoT). In IIoT systems, there is an increasing demand for different types of industrial equipment to exchange stream data with different delays. Communications between massive heterogeneous industrial devices and clouds will cause high latency and require high network bandwidth. The introduction of edge computing in the IIoT can address unacceptable processing latency and reduce the heavy link burden. However, the limited resources in edge computing servers are one of the difficulties in formulating communication scheduling and resource allocation strategies. In this article, we use deep reinforcement learning (DRL) to solve the scheduling problem in edge computing to improve the quality of services provided to users in IIoT applications. First, we propose a hierarchical scheduling model considering the central‐edge computing heterogeneous architecture. Then, according to the model characteristics, a deep intelligent scheduling algorithm (DISA) based on a double deep Q network (DDQN) framework is proposed to make scheduling decisions for communication. We compare DISA with other baseline solutions using various performance metrics. Simulation results show that the proposed algorithm is more effective than other baseline algorithms.
25 September 2019 in Optical Engineering
<!-- *** Custom HTML *** --> Virtualization is becoming a promising method for supporting various emerging applications, and achieving efficient embedding of virtual requests in elastic optical networks (EONs) has become a popular issue in network virtualization research. Meanwhile, multicast request mapping of EONs has become more challenging than that of conventional optical networks. Since the problem of minimizing the embedding cost and redundant transmission in multicast request embedding is NP-hard, we propose an efficient heuristic algorithm, called the virtual optical multicast-tree embedding algorithm based on the node correlation degree (VOME-NCD) to address the problem of low transmission redundancy. When embedding the virtual optical multicast request, it is based on the mapping of the correlation degree of each virtual node. Through extensive simulations, we demonstrate that VOME-NCD outperforms other embedding solutions in terms of the embedding cost, blocking probability, redundancy cost, and router hop count.
1 December 2021 in Optical Engineering
Network virtualization has been considered one of the promising paradigms for future network architecture, which can satisfy the diversified and personalized requirements of different applications by sharing physical network infrastructure. One of the major challenges in virtualization is how to map virtual networks efficiently onto the substrate network considering different constraints. We aim at solving the problem of high energy consumption of virtual network embedding in respect of the orthogonal frequency division multiplexing using elastic optical networks (EONs). First, we comprehensively consider and design an energy consumption model of network components, which denotes the key part of virtual optical network embedding (VONE). Then, a dynamic heuristic energy-efficient virtual optical network embedding (DEE-VONE) algorithm is proposed to achieve high energy efficiency and resource scalability. Simulation experiments have proven that, compared with the two benchmark algorithms, namely, the FFK-VONE and LRCK-VONE, the proposed method can maximize the network energy consumption saving at the cost of increasing the blocking probability. As the traffic load increases, the DEE-VONE method achieves higher energy-saving efficiency because optical switching energy saving plays an essential role.