安徽工业大学计算机科学与技术学院第三届网络与通信青年论坛邀请函

安徽工业大学计算机科学与技术学院

第三届网络与通信青年学者论坛

Video streaming plays a very important by enabling the on-line meeting, education and many other applications during the pandemic. However, the current on-line education/meeting/working applications are 2D video based, which do not allow users to select viewing angle, thus highly impeding the education and working efficiency. This problem can be solved by immersive video (e.g. VR/360 video, volumetric video), as it allows users to freely select any preferred viewing angle of the scene and thus provides them immersive viewing experience. This talk focuses on 6DoF video streaming. The challenges and the solutions are discussed.

报告主题： Resource allocation for 6DoF video streaming

当前以机器人为代表的智能产品（设备）越来越多地渗透到各领域，其智能感知能力是该新兴产业面临的重大难题，破解的关键是让机器人具有模拟人类多模态感知能力。本项研究聚焦复杂场景下人类细粒度行为与多维度情感感知这一老人看护、智能家居、智慧医疗等诸多领域的重大共性关键技术，旨在突破当前机械式人机交互掣肘、营造“自然”人机交互环境。拟推动无源感知与情感计算领域交叉，提出“无源行为感知-行为情感建模-多模态情感计算”研究体系，开拓无源感知驱动情感计算交叉方向，完成行为识别与情感计算方向的自然交叉和技术跨越：感知设备从有源到无源，感知空间从小局部到大范围、感知粒度由粗糙到细致；计算对象从个体到场景、计算维度从离散到连续、计算难度从简单到复杂，满足公共安全、智能家居、智慧医疗等诸多领域对“行为-情感透明感知空间”的迫切需求。

报告主题： WiFi-懂你的心跳与微笑

 A major disaster would damage the communication infrastructure severely, which results in further chaos and loss in the disaster stricken area. Rapid restoration of wireless/mobile communications is one of the most critical issues for disaster response. Wireless multi-hop networking by deploying low-cost relays is a promising solution to effectively extending the network services to people in the disrupted areas after largescale disasters occurred. In this talk, we discuss the following three great important issues in disaster response networking. First, how to accurately estimate the population distribution after the disaster. Second, how to judiciously place the limited number of relay nodes to maximize the population coverage ratio. Third, how to control the wireless access of user devices smartly and adaptively.

报告主题： Intelligent Networking for Post-disaster Scenarios

 In many real-time RFID-enabled applications (e.g., logistic tracking and warehouse controlling), a subset of wanted tags is often selected from a tag population for monitoring and querying purposes. How this subset of tags is rapidly selected, which is referred to as the subset selection problem, becomes pivotal for boosting the efficiency in RFID systems. Current state-of-the-art schemes result in high communication latencies, which are far from the optimum, and this degrades the system performance. This problem is addressed in this paper by using a simple Bit-Counting Function $BCF()$, which has also been employed widely by other protocols in RFID systems. In particular,  we first propose a near-OPTimal SeLection protocol, denoted by OPT-SL, to rapidly solve this problem based on the simple function $BCF()$. Second, we prove that the communication time of OPT-SL is near-optimal with rigorous theoretical analysis. Finally, we conduct extensive simulations to verify that the communication time of the proposed OPT-SL is not only near-optimal but also significantly less than that of benchmark protocols.

报告主题： A Near-optimal Protocol for the Subset Selection Problem in RFID Systems

王修君

报告主题： Collaborative Intelligence and Internet of Vehicles

 Future Internet of Vehicles systems feature a large number of devices and multi-access environments where different types of communication, computing, and storage resources must be efficiently utilized. It is therefore important to empower future IoV systems with advanced features, such as real-time reactive and proactive cooperation and coordination among different agents (or decision makers), including vehicles, roadside units, base stations, pedestrians, and other entities. Recently, artificial intelligence (AI) based approaches have been attracting great interest in empowering computer systems. In IoV systems, collaborative intelligence can be achieved via an efficient collaboration among heterogeneous entities, including vehicles, edges, and cloud. This talk will focus on the technical challenges and the synergistic effect of collaboration among heterogeneous entities and AI in enabling intelligent perception of environment, intelligent networking, and intelligent processing of big data in vehicular IoT systems.

报告主题： A Sample Efficient Model-based Deep Reinforcement Learning Algorithm with Experience Replay for Robot Manipulation

报告主题： A Distributed Learning Framework for Multi-Access Edge Computing in Beyond 5G Networks

 To improve the computation qualities of service and experience, multi-access edge computing (MEC) is envisioned as a promising paradigm by providing computing capabilities in close proximity to mobile users (MUs). In addition to local processing, a resource-constrained MU can also offload the computation to resource-rich MEC servers for remote execution. Beyond 5G networks are expected to enhance the 5G capabilities towards the support of seamless wireless connectivity. The trend of merging wireless communications and MEC motivates the investigation of computation offloading in beyond 5G networks. Nevertheless, the design of computation offloading policies remains challenging due to the environmental uncertainties and the limited resource sharing. In this talk, we adopt a multi-agent Markov decision process to formulate the computation offloading problems, for which a distributed learning framework is proposed. We examine the potentials of the proposed distributed learning framework through use case studies.