Research On Transactions Scheduling For Saving Energy Consumption In Mobile Real-time Database Systems

Recent advances in electronic technologies such as portable computers and wireless communication networks have led to the reality of distributed mobile computing systems, and users are able to access a collection of database services anytime using any computer(mobile or non-mobile)at any location. A lot of mobile computing systems, such as mobile electronical medical treatment systems, real-time transportation information management and navigation systems, and mobile Internet stock trading systems, are real-time essentially. The need of real-time data management systems for these mobile computing environments is strong. It is the basic, critical and core function of a real-time system that transactions are processing efficiently in order to satisfy the deadlines and energy consumption. The need to meet real-time time constraint while staying within energy constraint presents new design challenges.In mobile real-time database systems,transaction scheduling is an important aspect of transaction management. According to transaction's priority, transactions are processed. Traditional scheduling algorithms,such az Earliest Deadline First(EDF)Algorithm, Least Slack First(LSF)Algorithm etc. The priority is higher, the earilier is procedssed. energy-efficient scheduling is that frequency adjusted strategies are added to scheduling algorithms.real-time transactions's priority is not changed.Through adjusting frequency to a properly value which is based on the right time and right result, power consumption is saved.in the process of transaction,we can collect the slack time and dispatch some time to a transaction according to the length of transaction's slack time.so the execution time of the transaction is increased.The voltage of the CPU is decreased.according to the relationship between power consumption and voltage,the purpose about saving power consumption is reached.