Task Scheduling And Resource Allocation Optimization Approach In Cloud Platform For Large-scale Signal Deep Searching

With the development of military aircraft stealth technology,it is harder and harder to track the enemy stealth flight targets.The progress of the information age makes satellite traffic surges.The vast space between satellites and ground stations is filled with electromagnetic waves,and the passing stealth aircrafts will affect the space electromagnetic environment inevitably.By comparing the current electromagnetic wave signal with the historical electromagnetic wave signal,we may find the passing stealth aircrafts.A detection system like this will face the challenge of mass data calculation and the requirement of high real-time performance.Cloud computing technology provides a cost-effective and powerful platform for big data processing,it is well suited for military data analysis.Under this premise,we build a cloud platform for specific application,data is continually being transferred from multiple data sources to this platform for processing.Data manager can use geographically distributed computing centers to process their geographically dispersed data set.Usually,the data sets are dynamically generated and the resource cost varies over time,which makes it a critical issue of cost effectiveness to move the data from different geographic locations to different computing centers while providing suitable computation resources for processing.In this paper,we study a task scheduling and resource allocation optimization approach in cloud platform for large-scale signal deep searching.A pertinent joint stochastic optimization problem is firstly formulated,and then the problem is decoupled into two independent sub-problems with efficient solutions via Lyapunov framework.Next,an online algorithm based on the solutions is developed.Theoretical analysis shows that the proposed online algorithm can produce a solution which is arbitrarily close to the offline optimal solution while minimizing the data processing delays.Experiments on real dataset validate the proposed algorithms and demonstrate the superiority of the new approach.In addition,a large-scale signal deep searching cloud platform is built,and a stealth flight targets detection application is running on it.