The Research And Design For The Software Of Smart Home Management Based On Hadoop

With the continuous improvement of life quality by people in recent years,smart homes are increasingly favored by people.Most of the traditional smart home systems are independent with their own home server to provide services for user,but the redundancy,fault tolerance and data processing capabilities of the home servers are inefficient.The smart home under this mode is not flexible enough.In addition,the standards adopted by smart home manufacturers are not uniform,which makes it difficult for different manufacturers’ facilities to interconnect and communicate with each other.At the same time,it is also difficult to guarantee the security of underlying equipment data.Based on the above situation and the needs of users,By using the distributed computing platform Hadoop which has advantages of high redundancy,strong data processing ability etc.,combining Hadoop with the existing smart home system,and adopting the idea of hierarchical design,this study developes a loosely coupled and extensible smart home management software which is based on Hadoop,and provides users with a unified management services including flexible and extensible home system,which realizes interconnections and interoperability of intelligent equipment of different manufacturers and guarantees the security of underlying equipment data.The main work of this thesis is described as follow:1.Analyze the research status of the distributed computing framework technology and its application in smart home field.Based on the deficiencies of current smart home system,such as inadequate processing capacity of servers,the lack of interconnection between devices from different manufacturers etc.,taking into consideration the requirements of smart home system,the content and significance of this research are highlighted.2.Build the distributed computing platform Hadoop for this project.Meanwhile,considering that the software is prone to appear high concurrent access when providing services to users in multiple communities,this research also treats and improves the task scheduling algorithm for Hadoop platform.The Genetic-Ant Colony algorithm is improved by using multidimensional constraints goals and then apply the improved algorithm to Hadoop platform.According to the simulation results,compared with the Genetic-Ant Colony algorithm,the improved algorithm reduces the task execution time by 2.03% and improves the platform load balancing level by 1.44%,which improves user’s experience to a certain extent.3.According to the layered design idea and B/S design pattern,the smart home management software is divided into five layers: presentation layer,control layer,business logic layer,data access layer and data layer.By using Struts2,Spring4,Hibernate5(SSH)framework,meanwhile,according to the user’s requirements,this research design and implement various service functions of the software that include user login and registration,user management,data information management,equipment management,video surveillance,household appliances control and message push.4.According to the user’s demand for smart home services,the software quality test standards GB/T2500.51-2016,GB/T2500.10-2016 and the standard GB/T35136-2017,build a test environment for smart home management software to test the function and non-function of the software,which aims at testing the rationality and feasibility of its function.As the test results shows,the software satisfies the users’ needs in terms of smart home services,the software’s stability and compatibility also meet the national software standards.