With the development of artificial intelligence(AI)and 5G technology,the integration of sensing,communication and computing in the Internet of Vehicles(Io V)is becoming a trend.However,the large amount of data transmi...With the development of artificial intelligence(AI)and 5G technology,the integration of sensing,communication and computing in the Internet of Vehicles(Io V)is becoming a trend.However,the large amount of data transmission and the computing requirements of intelligent tasks lead to the complex resource management problems.In view of the above challenges,this paper proposes a tasks-oriented joint resource allocation scheme(TOJRAS)in the scenario of Io V.First,this paper proposes a system model with sensing,communication,and computing integration for multiple intelligent tasks with different requirements in the Io V.Secondly,joint resource allocation problems for real-time tasks and delay-tolerant tasks in the Io V are constructed respectively,including communication,computing and caching resources.Thirdly,a distributed deep Q-network(DDQN)based algorithm is proposed to solve the optimization problems,and the convergence and complexity of the algorithm are discussed.Finally,the experimental results based on real data sets verify the performance advantages of the proposed resource allocation scheme,compared to the existing ones.The exploration efficiency of our proposed DDQN-based algorithm is improved by at least about 5%,and our proposed resource allocation scheme improves the m AP performance by about 0.15 under resource constraints.展开更多
Task scheduling plays a crucial role in cloud computing and is a key factor determining cloud computing performance.To solve the task scheduling problem for remote sensing data processing in cloud computing,this paper...Task scheduling plays a crucial role in cloud computing and is a key factor determining cloud computing performance.To solve the task scheduling problem for remote sensing data processing in cloud computing,this paper proposes a workflow task scheduling algorithm—Workflow Task Scheduling Algorithm based on Deep Reinforcement Learning(WDRL).The remote sensing data process modeling is transformed into a directed acyclic graph scheduling problem.Then,the algorithm is designed by establishing a Markov decision model and adopting a fitness calculation method.Finally,combine the advantages of reinforcement learning and deep neural networks to minimize make-time for remote sensing data processes from experience.The experiment is based on the development of CloudSim and Python and compares the change of completion time in the process of remote sensing data.The results showthat compared with several traditionalmeta-heuristic scheduling algorithms,WDRL can effectively achieve the goal of optimizing task scheduling efficiency.展开更多
In this paper,we consider a cognitive radio(CR) system with a single secondary user(SU) and multiple licensed channels.The SU requests a fixed number of licensed channels and must sense the licensed channels one by on...In this paper,we consider a cognitive radio(CR) system with a single secondary user(SU) and multiple licensed channels.The SU requests a fixed number of licensed channels and must sense the licensed channels one by one before transmission.By leveraging prediction based on correlation between the licensed channels,we propose a novel spectrum sensing strategy,to decide which channel is the best choice to sense in order to reduce the sensing time overhead and further improve the SU's achievable throughput.Since the correlation coefficients between the licensed channels cannot be exactly known in advance,the spectrum sensing strategy is designed based on the model-free reinforcement learning(RL).The experimental results show that the proposed spectrum sensing strategy based on reinforcement learning converges and outperforms random sensing strategy in terms of long-term statistics.展开更多
现有移动群智感知中,大多研究将每个任务作为独立个体进行处理,对任务间约束关系缺乏研究,为此,提出了基于感知质量优先级的在线任务协作方法(online task collaboration method based on sensing quality priority,TCSP)。该方法首先...现有移动群智感知中,大多研究将每个任务作为独立个体进行处理,对任务间约束关系缺乏研究,为此,提出了基于感知质量优先级的在线任务协作方法(online task collaboration method based on sensing quality priority,TCSP)。该方法首先使用贪婪算法计算感知质量优先级,对全部任务进行筛选以保证任务完成率;然后将选出任务中存在时间先后或执行逻辑前后关系的多个子任务构建为任务协作图,并将其协作过程建模为有约束的马尔可夫决策过程,通过强化学习算法求出最优协作策略。实验结果表明,与现有基线方法相比,所提出的任务协作方法能够减少依赖任务的平均完成时间,有效降低平台的平均感知成本。展开更多
面向6G通信-感知-计算(通感算)融合的发展需求,亟需突破其资源高效分配算法。提出一种面向6G通感算融合的多粒度资源分配算法,该算法根据感知的网络状态以及基站自身状态,在多时间粒度上调整资源分配策略时间。首先,该算法将通信、感知...面向6G通信-感知-计算(通感算)融合的发展需求,亟需突破其资源高效分配算法。提出一种面向6G通感算融合的多粒度资源分配算法,该算法根据感知的网络状态以及基站自身状态,在多时间粒度上调整资源分配策略时间。首先,该算法将通信、感知、计算资源联合优化问题建模为多时间粒度上的最大化效用函数问题;其次,采用决斗深度Q网络(Dueling Deep Q-network, Dueling DQN)算法关注重要状态,忽视不重要状态,可以较快地找到最佳动作;最后在所搭建的实验平台中,将所提算法与多种资源分配算法进行对比,提高网络频谱效率,并降低传输时延、处理时延和资源分配动作执行的成本。展开更多
基金supported by The Fundamental Research Funds for the Central Universities(No.2021XD-A01-1)The National Natural Science Foundation of China(No.92067202)。
文摘With the development of artificial intelligence(AI)and 5G technology,the integration of sensing,communication and computing in the Internet of Vehicles(Io V)is becoming a trend.However,the large amount of data transmission and the computing requirements of intelligent tasks lead to the complex resource management problems.In view of the above challenges,this paper proposes a tasks-oriented joint resource allocation scheme(TOJRAS)in the scenario of Io V.First,this paper proposes a system model with sensing,communication,and computing integration for multiple intelligent tasks with different requirements in the Io V.Secondly,joint resource allocation problems for real-time tasks and delay-tolerant tasks in the Io V are constructed respectively,including communication,computing and caching resources.Thirdly,a distributed deep Q-network(DDQN)based algorithm is proposed to solve the optimization problems,and the convergence and complexity of the algorithm are discussed.Finally,the experimental results based on real data sets verify the performance advantages of the proposed resource allocation scheme,compared to the existing ones.The exploration efficiency of our proposed DDQN-based algorithm is improved by at least about 5%,and our proposed resource allocation scheme improves the m AP performance by about 0.15 under resource constraints.
基金funded in part by the Key Research and Promotion Projects of Henan Province under Grant Nos.212102210079,222102210052,222102210007,and 222102210062.
文摘Task scheduling plays a crucial role in cloud computing and is a key factor determining cloud computing performance.To solve the task scheduling problem for remote sensing data processing in cloud computing,this paper proposes a workflow task scheduling algorithm—Workflow Task Scheduling Algorithm based on Deep Reinforcement Learning(WDRL).The remote sensing data process modeling is transformed into a directed acyclic graph scheduling problem.Then,the algorithm is designed by establishing a Markov decision model and adopting a fitness calculation method.Finally,combine the advantages of reinforcement learning and deep neural networks to minimize make-time for remote sensing data processes from experience.The experiment is based on the development of CloudSim and Python and compares the change of completion time in the process of remote sensing data.The results showthat compared with several traditionalmeta-heuristic scheduling algorithms,WDRL can effectively achieve the goal of optimizing task scheduling efficiency.
基金supported by National Nature Science Foundation of China(NO.61372109)
文摘In this paper,we consider a cognitive radio(CR) system with a single secondary user(SU) and multiple licensed channels.The SU requests a fixed number of licensed channels and must sense the licensed channels one by one before transmission.By leveraging prediction based on correlation between the licensed channels,we propose a novel spectrum sensing strategy,to decide which channel is the best choice to sense in order to reduce the sensing time overhead and further improve the SU's achievable throughput.Since the correlation coefficients between the licensed channels cannot be exactly known in advance,the spectrum sensing strategy is designed based on the model-free reinforcement learning(RL).The experimental results show that the proposed spectrum sensing strategy based on reinforcement learning converges and outperforms random sensing strategy in terms of long-term statistics.
文摘现有移动群智感知中,大多研究将每个任务作为独立个体进行处理,对任务间约束关系缺乏研究,为此,提出了基于感知质量优先级的在线任务协作方法(online task collaboration method based on sensing quality priority,TCSP)。该方法首先使用贪婪算法计算感知质量优先级,对全部任务进行筛选以保证任务完成率;然后将选出任务中存在时间先后或执行逻辑前后关系的多个子任务构建为任务协作图,并将其协作过程建模为有约束的马尔可夫决策过程,通过强化学习算法求出最优协作策略。实验结果表明,与现有基线方法相比,所提出的任务协作方法能够减少依赖任务的平均完成时间,有效降低平台的平均感知成本。
文摘面向6G通信-感知-计算(通感算)融合的发展需求,亟需突破其资源高效分配算法。提出一种面向6G通感算融合的多粒度资源分配算法,该算法根据感知的网络状态以及基站自身状态,在多时间粒度上调整资源分配策略时间。首先,该算法将通信、感知、计算资源联合优化问题建模为多时间粒度上的最大化效用函数问题;其次,采用决斗深度Q网络(Dueling Deep Q-network, Dueling DQN)算法关注重要状态,忽视不重要状态,可以较快地找到最佳动作;最后在所搭建的实验平台中,将所提算法与多种资源分配算法进行对比,提高网络频谱效率,并降低传输时延、处理时延和资源分配动作执行的成本。
