HWHanfei Wen

Hanfei Wen

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Hanfei Wen has a total of 8 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • A Hybrid CNN-LSTM Architecture for High Accurate Edge-Assisted Bandwidth Prediction

    by Hanfei Wen, Jun Yu, Guangjin Pan, Xiaojing Chen, Shunqing Zhang, Shugong Xu

    In this letter, we propose an edge-assisted bandwidth prediction scheme based on a hybird convolutional neural network (CNN) and long short-term memory (LSTM) architecture. To accurately predict available downlink bandwidth with limited and regulated monitoring information, we decouple feature extraction and sequence learning to capture the feature vector by CNN and resolve the historical feature collection issue by LSTM, respectively. A two-stage training strategy is further adopted to improve the prediction accuracy and generalization ability. In addition, we deploy and evaluate the proposed bandwidth prediction scheme in our prototype system. The test results show that our proposed scheme improves the mean absolute error (MAE) and root mean square error (RMSE) by 81.4% and 82.0%, respectively, compared with the LSTM only prediction scheme, and consumes less than 3 ms running time, which is sufficient for many real-time video transmission applications.


  • Quality of Experience Oriented Adaptive Video Streaming for Edge Assisted Cellular Networks

    by Jun Yu, Hanfei Wen, Guangjin Pan, Shunqing Zhang, Xiaojing Chen, Shugong Xu

    HTTP adaptive streaming accounts for a large part of mobile Internet traffic. With the developing cellular communication technologies, the standard dynamic adaptive streaming over HTTP technique allows mobile terminals to adaptively configure the transmission rate of video streaming applications and achieve high quality of experiences (QoE) with many adaptive bitrate algorithms. However, existing schemes either neglect to consider radio access network (RAN) side conditions or only consider simple RAN information optimization schemes. In this paper, we propose a QoE-oriented adaptive video streaming scheme based on a dueling deep Q-learning network. The proposed scheme improves the QoE by jointly considering the physical layer transmission bandwidth and the higher layer buffer status. Through the numerical and prototyping results, we show that our proposed scheme outperforms the existing schemes, with the average QoE improvements of 12.6% to 28.8%.


  • Distributed Online Optimization of Edge Computing With Mixed Power Supply of Renewable Energy and Smart Grid

    by Xiaojing Chen, Hanfei Wen, Wei Ni, Shunqing Zhang, Xin Wang, Shugong Xu, Qingqi Pei

    Edge infrastructures, including edge computing servers, are increasingly powered by renewable energy and smart grid combined. Two-way energy trading allows the surplus or shortfall of renewable energy to be traded between a server and the smart grid, but is non-trivial due to randomly varying computation demands and renewables. This paper proposes a new online policy, namely, distributed online resource allocation and load management (DORL), which enables such an edge server and its serving devices to minimize their energy cost and energy consumption, respectively, in a fully distributed manner. The key idea is that we employ the stochastic dual-subgradient method to interpret the battery of the server as a virtual queue. Based on the virtual queue and task queues, the CPU frequencies of the devices and the edge server, the offloading transmit rates of the devices (to the server) and the energy trading decisions of the server (with the smart grid) are decoupled over time and among devices, and optimized on an ongoing basis. Furthermore, we prove that the DORL yields a feasible and asymptotically optimal solution with a cost-backlog tradeoff of $[\eta, 1/\eta]$ . Simulations show that the DORL reduces the system cost by nearly 50%, as compared to existing benchmarks.


  • Distributed Resource Allocation and Load Management for Smart Grid-Powered Edge Computing Systems

    by Xiaojing Chen, Hanfei Wen, Wei Ni, Shunqing Zhang, Xin Wang, Shugong Xu

    In this paper, we propose distributed online resource allocation and load management (DORL) for smart grid-powered edge computing (EC) systems. Taking into account renewable energy, two-way energy trading, and dynamic energy pricing, the resource allocation and load management task is formulated as an infinite-horizon optimization problem, which minimizes jointly the energy transaction cost of the base station (BS) and the energy consumption of the Internet-of- Things (IoT) devices. Building on the stochastic dual-subgradient method, the proposed DORL algorithm decouples the real-time decisions of the BS and devices, and stabilizes the backlogs of computation tasks with a cost-backlog tradeoff of $[\eta, \frac{1}{\eta}]$ , for any $\eta > 0$ . Theoretical analyses prove that the proposed DORL can yield a feasible and asymptotically optimal solution.