Research On Edge Server Placement And Service Placement And Sensing Task Allocation In Mobile Computing

The emergence of 4G/5G has driven the development of mobile edge computing(MEC)and mobile crowd sensing(MCS),among which edge server placement,service placement and sensing task allocation have attracted extensive attention from the academic community.In this article,we study edge server deployment and service placement of MEC,and task allocation of MCS,utilizing clustering algorithm,nonlinear programming,submodular optimization theory and so on.The main work includes:(1)A two-step method for joint edge server placement and service placement is proposed,taking the economic benefit of MEC platform into consideration.There are rare works considering edge server placement and service placement simultaneously.However,different edge server placement schemes will lead to different services that need to be placed for each edge server and affect subsequent service placement for each edge server.In addition,due to different service request rates and prices,appropriate service placement solutions are needed to increase the overall profit of MEC system.In this paper,we design a joint edge server deployment and service placement model with the goal of maximizing the overall profit of all edge servers under the constraints of the number of edge servers,the relationship among edge servers and base stations,the storage capacity and the computing capacity of each edge server.Also,the service placement explicitly takes into account the structure of current edge server placement and different service request rates and prices.Then we propose a two-step method including the clustering algorithm,nearest neighbor algorithm and interior-point algorithm to solve the formulated problem.Extensive evaluations based on the real-world dataset demonstrate that the proposed algorithm outperforms the baseline methods.(2)A duration-sensitive task allocation method which meets the requirement of sensing duration for MCS tasks is proposed.There are few studies focusing on allocating sensing tasks with specific sensing duration.However,sensing duration plays a key role for the success of many sensing tasks.For example,when the crowd sensing system needs to monitor the crowd flow in locations of interest,it is better to allocate this task to workers who can record a video of certain duration rather than those who can only take a picture.In this article,we design a duration-sensitive task allocation model by considering the requirement of sensing duration explicitly.The model aims at maximizing the number of completed tasks under the constraints of sensing duration and task capacity of each worker.We design a utility function that can reflect the probability of task completion by using the exponential distribution and transform the formulated problem to a constrained submodular optimization problem.Then,according to the submodular optimization theory,an efficient greedy heuristic for task assignment is proposed based on the utility function and the approximation ratio of the proposed algorithm is 1-1/e.Extensive evaluations based on the simulated and real-world datasets demonstrate that the proposed algorithm outperforms the baseline methods.