Seismic imaging of Northwest Pacific and East Asia: New insight into volcanism, seismogenesis and geodynamics

We examine the upper mantle and transition zone beneath the western Pacific using multiple ScS reverberations. A low-velocity zone (LVZ) is found above the 410-km discontinuity oceanward of the subducting Honshu slab at an average depth of 356 km, with a thickness that ranges from 50 to 75 km assuming the LVZ continues to the 410-km discontinuity, which is locally elevated. The low-velocity region is evident in previous tomographic studies, and our results suggest that the anomaly is best explained by a layer of partial melt. The layer may be entrained from above by subduction or produced in situ by the combined effects of water and temperature. A self-consistent model that explains local P-wave velocities (Obayashi, M., Sugioka, H., Yoshimitsu, J., Fukao, Y., 2006. High temperature anomalies oceanward of subducting slabs at the 410-km discontinuity. Earth and Planetary Science Letters 243, 149–158) and our observations calls for a maximum temperature anomaly of ∼150 °C and a resulting maximum olivine water content after melting of 0.100 wt. %.

Melting in the Deep Upper Mantle Oceanward of the Honshu Slab

A new P wave tomographic model of the mantle was constructed using more than 10 million travel times. The finite-frequency effect of seismic rays was taken into account by calculating banana-donut kernels at 2 Hz for all first arrival time data, and at 0.1 Hz for broadband differential travel time data. Based on this model, a systematic survey for subducted slab images was developed for the circum-Pacific; including the Kurile, Honshu, Izu-Bonin, Mariana, Java, Tonga-Kermadec, southern and northern South America, and Central America, arcs. This survey revealed a progressive lateral variation of the configuration of slabs along arc(s), which we interpret as an indication for successive stages of slab subduction through the Bullen's transition region with the 660 km discontinuity at the middle. We identified the four distinct stages: I - slab stagnant above the 660 km discontinuity; II - slab penetrating the 660 km discontinuity; III - slab trapped in the uppermost lower mantle (at a depth of 660–1000 km); and IV - slab descending well into the deep lower mantle. The majority of slab images are found to be either at Stage I or III, suggesting that Stages I and III are relatively stable or neutral and II and IV are relatively unstable or transient. There is a remarkable distinction for the deepest hypocentral distribution between slabs at Stage I and slabs at Stages II or III.

A finite-frequency P-wave tomographic model: images of subducted slabs stagnant above, penetrating through and trapped below the 660-km discontinuity

B–Sr–Nd–Pb isotopic constraints on the origin of the Maoniuping alkaline syenite–carbonatite complex, SW China

Waveforms of SS precursors recorded by global stations are analyzed to investigate lateral heterogeneities of upper-mantle discontinuities on a global scale. A sporadic low-velocity layer immediately above the 410-km discontinuity (LVL-410) is observed worldwide, including East Asia, western North America, eastern South America, the Pacific Ocean, and possibly the Indian Ocean. Our best data coverage is for the Pacific Ocean, where the LVL-410 covers 33–50% of the resolved region. Lateral variations of our LVL-410 observations show no geographical correlation with 410-km discontinuity topography or tomographic models of seismic velocity, suggesting that the LVL-410 is not caused by regional thermal anomalies. We interpret the LVL-410 as partial melting due to dehydration of ascending mantle across the 410-km discontinuity, which is predicted by the transition zone water filter hypothesis. Given the low vertical resolution of SS precursors, it is possible that the regions without a clear LVL-410 detection also have a thin layer. Therefore, the strong lateral heterogeneity of the LVL-410 in our observations suggests partial melting with varying intensities across the Pacific, and further provides indirect evidence of a hydrous mantle transition zone with laterally varying water content.

A sporadic low-velocity layer atop the 410-km discontinuity beneath the Pacific Ocean

Pervasive low-velocity layer atop the 410-km discontinuity beneath the northwest Pacific subduction zone: Implications for rheology and geodynamics

Compared with traditional cloud computing schemes, edge computing provides an effective solution towards meeting real-time, security and reliability demands of compute-intensive and delay-sensitive Industrial Internet of Things (IIoT) applications. However, many existing studies only optimize end-edge-cloud cooperative task offloading, and neglect the optimization of the communication and computation resource allocation. In this paper, the cooperative partial task offloading and resource allocation (CPTORA) framework is designed, which jointly considers cooperation among various IIoT devices, local edge computing servers (ECSs), non-local ECSs, and cloud computing servers. Then, considering the complex dynamics and unpredictability in IIoT environments, the joint optimization problem is modeled as a constrained Markov decision process. Furthermore, we propose an improved soft actor-critic-based CPTORA (ISAC-CPTORA) algorithm, to make task offloading and resource allocation decisions for each IIoT device. This algorithm innovatively introduces the idea of distributional reinforcement learning to the soft actor-critic, which can effectively reduce Q-value overestimations or underestimations. Meanwhile, this algorithm employs the prioritized experience replay to enhance its learning efficiency. Extensive laboratory experiments indicate that our CPTORA framework and ISAC-CPTORA algorithm efficiently decrease the total system costs, in contrast to various baseline frameworks and algorithms. 

Deep Reinforcement Learning Based Cooperative Partial Task Offloading and Resource Allocation for IIoT applications

Obstructive sleep apnea (OSA) syndrome is a common sleep disorder and a key cause of cardiovascular and cerebrovascular diseases that seriously affect the lives and health of people. The development of Internet of Medical Things (IoMT) has enabled the remote diagnosis of OSA. The physiological signals of human sleep are sent to the cloud or medical facilities through Internet of Things, after which diagnostic models are employed for OSA detection. In order to improve the detection accuracy of OSA, in this study, a novel OSA detection system based on manually generated features and utilizing a parallel heterogeneous deep learning model in the context of IoMT is proposed, and the accuracy of the proposed diagnostic model is investigated. The OSA recognition scheme used in our model is based on short-term heart rate variability (HRV) signals extracted from ECG signals. First, the HRV signals and the linear and nonlinear features of HRV are combined into a one-dimensional (1-D) sequence. Simultaneously, a two-dimensional (2-D) HRV time-frequency spectrum image is obtained. The 1-D data sequences and 2-D images are coded in different branches of the proposed deep learning network for OSA diagnosis. To validate the performance of the proposed scheme, the Physionet Apnea–ECG public database is used. The proposed scheme outperforms the existing methods in terms of accuracy and provides a novel direction for OSA recognition.

Obstructive Sleep Apnea Detection Scheme Based on Manually Generated Features and Parallel Heterogeneous Deep Learning Model Under IoMT

As a prototype of the underwater Internet of Things-enabled maritime transportation systems, multi-Autonomous Underwater Vehicle (AUV)-based Underwater Wireless Networks (UWNs) have become an important research topic due to their distribution and robustness. In this paper, the concept of multi-AUV-based UWNs is first defined, where AUV is regarded as a network node, and communication among the AUVs is the potential network links. Then, to improve network scalability and controllability, a paradigm of Software Defined multi-AUV-based UWNs (SD-UWNs) is proposed, where the Software Defined Network (SDN) technique is used to upgrade the UWN architecture by directing intelligent network functions. Topology and artificial potential field theories are applied to construct a network control model for the SD-UWNs. Based on the efficient data sharing ability of the SD-UWNs, an early warning obstacle avoidance-enabled path planning scheme is proposed to guarantee safe sailing of the SD-UWNs, where comprehensive obstacle avoidance scenarios are taken into account. Simulation results demonstrate that the proposed method is effective in planning the cooperative operation for the SD-UWNs and is capable of performing accurate and reliable obstacle avoidance tasks.

Early Warning Obstacle Avoidance-Enabled Path Planning for Multi-AUV-Based Maritime Transportation Systems

The exploitation/utilization of marine resources and the rapid development of urbanization along coastal cities result in serious marine pollution, especially underwater diffusion pollution. It is a non-trivial task to detect the source of diffusion pollution, such that the disadvantageous effect of the pollution can be reduced. With the vision of 6G framework, we employ Autonomous Underwater Vehicle (AUV) flock and introduce the concept of AUV-based network. In particular, we utilize the Software-Defined Networking (SDN) technique to update the controllability of the AUV-based network, leading to the paradigm of SDN-enabled multi-AUVs network Intelligent Transportation Systems (SDNA-ITS). For SDNA-ITS, we utilize artificial potential field theories to model the control model. To optimize the system output, we introduce the graph-based Soft Actor-Critic (SAC) algorithm, i.e., a category of Multi-Agent Reinforcement Learning (MARL) mechanism where each AUV can be regarded as a node in a graph. In particular, we improve the optimization model based on Centralized Training Decentralized Execution (CTDE) architecture with the assistance of the SDN controller, by which each AUV can efficiently adjust its speed towards the diffusion source. Further, to achieve exact path planning for detecting the diffusion source, a dynamic detection scheme is proposed to output the united control policy to schedule the SDNA-ITS dynamically. Simulation results demonstrate that our approaches are available to detect the underwater diffusion source when the actual scenario is taken into account and perform better than some recent research products.

Guangjie Han

Papers

Pervasive low-velocity layer atop the 410-km discontinuity beneath the northwest Pacific subduction zone: Implications for rheology and geodynamics

Deep Reinforcement Learning Based Cooperative Partial Task Offloading and Resource Allocation for IIoT applications

Obstructive Sleep Apnea Detection Scheme Based on Manually Generated Features and Parallel Heterogeneous Deep Learning Model Under IoMT

Early Warning Obstacle Avoidance-Enabled Path Planning for Multi-AUV-Based Maritime Transportation Systems

Smart Underwater Pollution Detection Based on Graph-Based Multi-Agent Reinforcement Learning Towards AUV-Based Network ITS