With the development of mobile cloud computing and the improvement of users’ performance requirements for mobile terminals, transferring the complex computing and storage requirements of mobile terminals to the cloud for processing is an inevitable trend of mobile cloud computing. Therefore, how to reduce the energy consumption of mobile terminals and clouds , Improving energy utilization and enhancing user experience are one of the key issues that green cloud computing urgently needs to solve. This paper mainly studies the energy consumption optimization problem in the data transmission process of mobile cloud computing and proposes a method to minimize unit data based on optimal stopping theory. The average energy consumption transmission strategy; the optimal transmission strategy based on the secretary problem. By constructing a data transmission queue model with multiple applications, in the secretary problem based on the minimum absolute ranking mean of the selected candidates, it is proposed to let k candidates go After seeing the rules of excellence, it is proved that there is an optimal k value for this rule. The simulation results show that the optimization strategy proposed in this article has smaller average energy consumption per unit of data, better energy consumption efficiency and better detection efficiency. .
With the rapid development and widespread application of mobile devices, mobile terminals such as mobile phones and PADs have become the main devices for people to chat, entertain and work. However, mobile terminals are limited by their size and energy, and always have weak computing power and storage space. Various problems such as small size and short battery life [1. Therefore, in order to make up for the various shortcomings of mobile terminals, transferring complex computing and storage requirements to the cloud for processing is an inevitable trend in mobile cloud computing. According to statistics, as of 2017 , mobile cloud traffic already accounts for 84% of the entire mobile traffic. And according to IDC’s prediction, the total global mobile data volume will reach 40,000 EB in 2020, with a compound annual growth rate of 36%, and mobile cloud traffic will account for 94% of the entire mobile traffic. ; And China Internet
The growth rate of Internet data traffic is even more prominent. In 2020, China’s Internet data traffic will reach 8806 EB, accounting for 22% of global data production, with a compound annual growth rate of 49%. This puts forward new requirements for the efficiency of data transmission between mobile terminals and the cloud. A huge challenge. A better data transmission strategy can save the energy consumption of mobile terminals, optimize and improve the performance of mobile terminals. Therefore, this paper optimizes the energy consumption of mobile terminals and cloud data transmission in mobile cloud computing. , an optimization strategy based on optimal stopping theory is proposed to reduce the energy consumption of mobile terminals, improve energy utilization, increase user experience, and lay a solid foundation for promoting the development of mobile cloud computing.
The data transmission channel of mobile cloud computing is the same as the traditional Internet wireless channel, which is updated in real time with changes in time and space [2]. If the mobile terminal transmission power
Remaining unchanged, the greater the transmission rate, the greater the amount of data transmitted in the same time, and the smaller the average energy consumption per unit of data. Furthermore, if the mobile terminal continuously observes the channel conditions of the wireless link, then select Sending data at a time of good quality can greatly reduce the energy consumption of data transmission.
In fact, the problem of mobile terminals selecting times with better channel conditions to transmit data is a distributed opportunistic scheduling problem. 3. We can use the optimal stopping rule to solve this problem. The optimal stopping rule is a decision-maker based on continuously observed random variables. With the goal of maximizing rewards or minimizing expected costs, it is decided to choose a suitable moment to take a given behavior. In the research of scholars, the optimal stopping theory has solved many optimization problems. For example, the literature [4] adopts the optimal stopping theory. Optimal stopping theory studies the optimal scheduling problem of information consumption in self-organizing networks. Literature [5] uses optimal stopping theory in mobile networks to study the energy consumption optimization problem of multiple sending terminals using the same channel for data distribution. Literature [6] The optimal stopping theory is used to obtain the optimal expected reward of the optimal relay node, thereby achieving the optimal energy-efficient routing strategy.
In the optimal stopping rule problem, our goal is to find in the cost function
A stopping rule when the transmission rate is maximum, and this algorithm is used to minimize the average energy consumption per unit of data. This article mainly studies mobile cloud computing
For the energy consumption optimization problem in the data transmission process, a data transmission energy consumption optimization strategy based on optimal stopping rules is proposed. The specific research ideas are as follows: under a given data generation rate, by constructing a data transmission queue model with multiple applications , In the secretarial problem based on the minimum absolute ranking mean of the selected candidates, a rule is proposed to let k candidates go and then accept the best candidates. The energy consumption and delay in the transmission process are comprehensively considered to minimize the unit data. average energy consumption.
The organization structure of this paper is as follows: Section 2 introduces related research work; Section 3 explains the system model and related theories; Section 4 discusses the optimization strategy for minimizing expected energy consumption based on optimal stopping theory; Section 5 conducts simulation experiments and analysis Experimental results; finally Section 6 summarizes the full text and discusses the next research work.
Related research work
In recent years, how to reduce the energy consumption of mobile terminals and clouds during the data transmission process of mobile cloud computing has attracted a lot of research by scholars. Researchers mainly focus on two directions: first, optimizing data models in mobile terminals and clouds to reduce energy consumption; The second is to optimize the algorithm to reduce energy consumption during the data transmission process.
In the study of mobile terminal and cloud energy consumption optimization, literature [2] constructed a cloud-assisted mobile platform. Based on this platform, mobile applications can be executed on the mobile terminal or in the cloud. When executing the program on the mobile terminal , optimize energy consumption by adjusting the CPU frequency; when executing programs in the cloud, minimize energy consumption by optimizing the data transmission rate. Literature [7] proposes a collaborative mobile cloud system. The system forms multiple UE (User Equipment) alliances And receive part of the data requested from the base station, and then exchange the received data with each other, which greatly saves time cost and energy consumption. Literature [8] proposes a joint collaboration and channel selection framework for mobile terminal data offloading, which is executed in a cooperative manner. The data is offloaded to the cloud; and a distributed channel selection algorithm is designed using the Markov approximation method, so that each mobile terminal can organize itself into a stable structure without the exchange of information across the entire network, and only realize data between mobile devices on the same channel. Exchange. Literature [9] proposes a joint offloading model. Mobile devices can perform tasks locally, offload tasks to other mobile devices or directly transmit tasks to the cloud for execution according to the allocation of the model controller; and the fingerprint technology is used to describe each task. Similarity, mobile devices can share computing resources for similar tasks with each other to achieve slowdown
The purpose of solving data traffic pressure and reducing energy consumption of mobile terminals. Literature [10] performs joint collaboration between energy-poor and data-poor mobile devices to share resources. When mobile devices offload computing tasks to the cloud, energy-poor and energy-poor The mobile device of the collection module collects energy from the data-poor mobile device through the energy division strategy, and the energy-poor mobile device will also help the data-poor mobile device to perform calculations.
In the study of energy consumption optimization in the data transmission process of mobile cloud computing, literature [11] proposed an application layer adaptive transmission protocol. This protocol uses a stochastic optimization framework to decide through a low-complexity and low-overhead online algorithm. Whether the wireless connection is efficient. This article does not set a threshold to judge whether the wireless connection is efficient, but makes transmission decisions adaptively in achieving multi-application energy delay balance. Literature [12] proposes a novel data transmission Optimization algorithm: DTM algorithm. This article adds a proxy server layer between the cloud and the mobile terminal. This layer is responsible for processing requests from the mobile terminal and responses from the cloud; the DTM algorithm uses parallel
Processing and caching technology optimizes transmission tasks and helps reduce the energy consumption of mobile devices and user waiting time while ensuring the quality of data transmission. Literature
[13] proposed a novel online prefetching technology, which performs prediction and prefetching simultaneously within a time period, predicts data in real time, and avoids prefetching a large amount of unnecessary data. Literature [14] is based on single-user MCO (Mobile computation offloading) system proposes an online prefetching technology. This technology seamlessly integrates task-level calculation prediction and real-time prefetching when the program is running. When the possibility of data being executed exceeds the threshold, it Prefetch the data to be used, and the scale of data prefetching increases linearly with the possibility of the data being executed. Based on Lyapunov optimization, the literature [15] proposed a new efficient method in the process of data transmission between mobile terminals and clouds. A powerful data transmission strategy. This strategy actively and adaptively captures frequently used data when the channel state is good, while satisfying the time delay, to avoid unnecessary energy consumption caused by retrieving data when the channel state is poor. Literature [16] Compromise optimization of energy consumption and delay from the user’s perspective. This paper considers three QoE (Quality of Experience) domains, derives a QoE-aware cost function including energy and delay, and proposes an efficient approximate dynamic program Optimization algorithm. Literature [17] uses queuing theory to deeply study the joint optimization of energy consumption, delay and payment cost in the data offloading process in fog computing systems. Different queue models are established for mobile terminals, fog terminals and cloud terminals respectively. Using the scalarization method, the multi-objective optimization problem is converted into a single-objective optimization problem, thereby achieving the purpose of reducing the energy consumption of mobile terminals.
In summary, although researchers have optimized the energy consumption of mobile terminals and clouds in mobile cloud computing and the data transmission process, they have not comprehensively considered factors such as dynamic data arrival and delay. Therefore, this article is based on the optimal stopping theory , comprehensively considering factors such as energy consumption, bandwidth and delay, and minimizing the average energy consumption per unit of data transmission under the premise of dynamic arrival of data.
Keywords: Internet of Things gateway