Investigation On Efficient Resource Allocation Of Intelligent Edge For Internet Of Things

With the rapid development of the Internet of Things(Io T),5G technology and Artificial Intelligence(AI),series of new intelligent edge applications based on the Internet of Things(for example,smart city,intelligent security,unmanned intelligent driving,etc.)have emerged as the promising directions in the future.However,in the face of a large amount of random distributed and hugh energy consumed intelligent edge devices,how to exploit the joint communication and computation design for improving the system energy efficiency while ensuring that model training performance training speed and the model accuracy has become a critical issue needed to be resolved.Based on the above considerations,this paper firstly focus on the joint optimization design of communication and computing resources allocation under the federated edge learning framework.From the perspective of the most intuitive point of view,namely,energy efficiency optimization,the battery life of intelligent edge terminal equipment is improved by reducing the energy consumption of intelligent edge.Secondly,facing complex edge intelligence scenario(such as industrial automation,intelligent agriculture,etc.),large scale of energy consumption and some application requirement of a quick arrangement of the new power cost-effective way under new intelligent edge scenario,it is difficult to solve the problem of energy crisis while only considering energy consumption optimization.Based on the above considerations,this paper aims to study the efficient UAV-enabled wireless power transfer,aiming to achieve the efficient resource allocation design for the intelligent edge of the Internet of Things from the two aspects of energy efficiency and energy supply.The main work of this paper has been devided into two aspects,which are shown as follows:(1)We consider the energy efficienct allocation design of edge terminal devices under future Io T scenario,especially the emerging federated edge learning framework.This paper aims to design a joint communication and computation optimization scheme by optimizing the computing and communication resources allocation.We propose efficient algorithms to solve the formulated energy minimization problems by using the techniques from convex optimization.Numerical results show that as compared to other benchmark schemes,our proposed joint communication and computation design significantly improves the energy efficiency of the federated edge learning system,by properly balancing the energy tradeoff between communication and computation.(2)Based on the fairness principle,we consider the efficient energy transmission strategy which is applicable to the future Io T edge network,we proposed the hovering optimization by taking advantage of the convenience of UAV deployment as well as the high efficiency of wireless energy deployment,aiming to maximize the minimum received wireless energy tranfer of intelligent edge terminals.In this paper,we innovatively employed the radio map technology to precisely describe the whole wireless channel information for better wireless energy transfer performance.Numerical results also show that our proposed algorithm significantly improves the WPT performance,as compared with conventional designs based on Lo S and probabilistic Lo S channel models.