| With the continuous development of Internet technology, a variety of Internet and medium-sized enterprises are faced with increasingly large business data. In the face of such massive amounts of data, and how through the use of efficient information handling and data storage mode, solve application is running a huge amount of computation speed and the conflict between the software industry has became a hot issue.The graduation design choice a location-based service software which run on the Android platform as a background, use Hadoop and HBase as massive data carriers, and through clever design of the Java interface, the distribution will be calculated by a large number of nodes to a cluster composed,in order to provide research this contradiction between ultra-large-scale mass data computing services and system efficiency point.This thesis begin with the introduction of basic theory of the cloud computing platform Hadoop and non-relational database----HBase, analyzes the system performance requirements of front-end program, followed by a detailed description of the system in HBase logical model and physical storage model design and implementation, and finally on the basis of Hadoop is designed based on a system functional interface. After the initial system features modular design and testing verified the Hadoop system feasibility study for the project indicates a clear set of next research direction. |