Design And Implementation Of Message Distribution System Based On Integrated Data Service Platform

Recently, with the rapid development of the mobile communications network and smart handhold devices, the method and demand of communication have changed. The traditional voice communication is no longer sufficient to meet all the needs of the users. Besides, with the rise of the mobile Internet and social dramatic increase in the amount of information, people want to be able to obtain a variety of information anytime, anywhere. All of these provide the foundation for the rise of mobile data services. In this environment, a precise marketing business is becoming more and more popular. The business is on the basis of user data mining to identify target groups. Businesses through SMS(Short Message Service), MMS(Multimedia Messaging Service) and Wappush to push the information to the user’s mobile terminal. Due to the large number of targets and the requirements of intelligent delivery, we need to establish an intelligent, high-performance platform to handle this type of message distribution.The focuses of the paper mainly include:1) Design and implementation of load balancing strategy to ensure that the message is accurate, timely sent.2) Design and implementation of the flow control strategy during the message distribution process to adapt to the complex network environment.3) Design and implementation of the priority control program to respond to different business needs.4) Design and Implementation of disaster recovery and alarm function to improve disaster recovery and fault tolerance of the system. In the part of load balancing, this paper introduces a dynamic load balancing decision-making model. On the basis of detailed analysis of the factors that affect the application server load, this paper presents a load balancing decision model. By derivation and simplification, this paper provides a full load balancing decisions model. In the part of Priority control, the system assigns different transmission rates for different priority task, so that high priority task can take up more bandwidth resources. In aspect of flow control, the paper introduces three methods, including number disordering, multi-processing channel and dynamic rate adjustment. The number disordering is designed to prevent the issue of concentration of the numbers belonging to a certain base station. The multi-processing channel is designed to adapt to the complex network environment. Dynamic rate adjustment is introduced in order to adapt to the real-time changes of the network environment, so that the rate of message sending is able to adjust the current network load dynamically. In the part of disaster recovery and alarm, the disaster recovery program includes the data disaster recovery and the application disaster recovery. In terms of data disaster recovery, the introduction of database hot standby and counting server to prevent data loss, and to provide a basis for continued service capabilities of the system. The application disaster recovery includes exception detection and taking over the task of the abnormal server.