This paper addresses optimal allocation and sizing of inverter-based distributed Generation (DG) with the aim of reducing power loss, enhancing voltage profile and preserving power quality of radial distribution netwo...
详细信息
distributed training crossing multiple computing nodes and accelerators has been the mainstream solution for large model training. Precedent work on distributed deep learning (DDL) training acceleration has focused on...
详细信息
Phase retrieval is a crucial step in processing data from advanced X-ray diffraction imaging experiments to analyze the 3D structure of biological macromolecules. However, when the 3D volume is large-scale and consist...
详细信息
In this paper, the problem of sensor fault detection and isolation (FDI) for neutral time-delay systems with uncertain disturbances is studied. The sensor fault is transformed into actuator fault by state space equati...
详细信息
With the continuous advancement of large-scale models and expanding volumes of data, a single acceleration hardware is no longer sufficient to meet the training demands. Simply stacking multiple acceleration hardware ...
详细信息
Photonic integrated circuits provide compact and efficient solutions for multimodal spectroscopic sensors. However, the resulting sensory data is highly complex and contains significant redundancy. To circumvent high ...
详细信息
In this paper, we investigate the power allocation problem to maximize the long-term average downlink sum-rate for orthogonal frequency division multiplexing (OFDM) based multi-cell networks. The traditional centraliz...
详细信息
ISBN:
(纸本)9798350333398
In this paper, we investigate the power allocation problem to maximize the long-term average downlink sum-rate for orthogonal frequency division multiplexing (OFDM) based multi-cell networks. The traditional centralized training and centralized execution (CTCE) scheme is impractical to solve this complex problem, because it is hard for training center to obtain the global information, and the signaling overhead is also unbearable. To this end, a centralized training and distributed execution (CTDE) framework for wireless power control based on deep reinforcement learning (DRL) is proposed. Specifically, we first transform this problem into a sequential decision problem by designing appropriate state, action and reward. Then, a CTDE scheme, where each agent only requires its own local observations for power allocation, is devised through combining QMIX algorithm and Normalized Advantage Functions (NAF) algorithm. In particular, QMIX network is used to fit the total Q-value function and NAF network is deployed to handle continuous power output. Simulation results show that the proposed scheme converges fast and achieves better rate performance compared with conventional CTDE algorithms.
In this work, our aim is to determine the viability (feasibility) of quantum task mapping algorithms on distributed heterogeneous computingsystems. To quantify viability, we compared the quantum algorithms with nine ...
详细信息
Service placement in cloud-edge environments is complex because workloads must be placed on constrained nodes based on particular objectives, like response time, energy, or cost. Many advanced techniques emerged over ...
详细信息
ISBN:
(纸本)9798331539580
Service placement in cloud-edge environments is complex because workloads must be placed on constrained nodes based on particular objectives, like response time, energy, or cost. Many advanced techniques emerged over time to tackle this issue. However, real-world experiments are the minority. Theoretical and simulation-based evaluations are prevalent. We present a Platform for Universal and Lightweight Cloud-Edge Orchestration (PULCEO) to foster real-world evaluations. It supports creating, operating, monitoring, evaluating, and documenting orchestration solutions via a RESTful API. For evaluation, we performed a case study. We used PULCEO to reproduce a representative and theoretically designed solution for service placement in a real-world environment. Our platform can transfer theoretical orchestration solutions to real-world environments. Consequently, our platform simplifies realworld experiments with topology creation, dynamic link quality measurement, evaluation, and documentation automation.
Indoor positioning systems (IPS) continue to encounter significant challenges in achieving meter-level accuracy, particularly in large and intricate environments such as airports, hospitals, and industrial sites. Desp...
详细信息
暂无评论