Abstract—Real-time disk scheduling (RTDS) plays an important role in time-critical applications. The main idea in real time database system is that the correctness of the system depends not only on the logical results of the computations but also on the time at which the results are produced. Due to rigorous timing requirements for error free output, data must be accessed under real-time constraints. Therefore how to maximize data throughput under real-time constraints poses a big challenge in the design of real-time disk scheduling algorithms. Numbers of algorithms are proposed to schedule real time transactions in order to increase the overall performance. Currently Earliest-Deadline-First (EDF) is a basic algorithm which meets the real time constraints, but it gives poor disk throughput. Scan-EDF work only for those transactions which are having same deadline. In 2006 g-EDF algorithm has been proposed which works after making groups for transaction having close deadlines. In groups it apply SJF algorithm. We also propose a new algorithm “FEASIBLE GROUP-EDF” that works both in under load and overload conditions as well as show better throughput than earliest mentioned algorithms. It also makes groups and applies SSTF algorithm as well as check feasibility of transaction. Index Terms—EDF, SCAN-EDF, G-EDF, SJF, SSTF, real- time, overloaded. I. INTRODUCTION Traditionally, real-time system manages their data in application dependent structures. As real-time systems evolve, their applications become more complex and require accessing more data. It thus becomes necessary to manage the data in more systematic and organized manner. Database management system provides tools for such organization, so in recent year there has been interest in “merging” database and real-time system. The resulting integrated system which provides the database operations with real-time constraint is called as real-time database system (RTDBS) [1], [2]. A Real-Time Database System (RTDBS) is a transaction processing system that is designed to handle transactions with the timing constraint. Several previous RTDBS as in studies had been done to address the issue of scheduling transactions with the objective of minimizing the number of miss transaction. A common observation of these studies has been that, if we assigning priorities to transactions according to an Earliest. Deadline policy minimizes the number of miss Manuscript received February 14, 2013; revised April 14, 2013. S. Y. Amdani is with Department of CSE, BNCOE, Pusad, India (e- mail: salimamdani@yahoo.com). M. S. Ali is with P. R. M. C. E. & M, Badnera, India (e-mail: softalis@gmail.com). transactions in systems operating under low or moderate levels of workload condition. This is due to earliest Deadline giving the highest priority to transactions that have the least remaining time in which to complete. These studies have also observed that the performance of Earliest Deadline steeply degrades in an overloaded system. This is because, under heavy loaded workload condition transactions gain high priority only when they are close to their deadlines. While investigating scheduling algorithms, we have analyzed a variation of EDF that can improve success ratios, particularly in overloaded conditions. The algorithm can also decrease the average response time for tasks that is group-EDF, or gEDF, where the tasks with “similar” deadlines are grouped together (i.e., deadlines that are very close to one another), and the Shortest Job First (SJF) algorithm is used for scheduling tasks within a group. It should be noted that our approach is different from adaptive schemes that switch between different scheduling strategies based on system load; Fg-EDF is used in overloaded as well as underloaded conditions. II. BACKGROUND A. EDF Algorithm The idea of EDF was published in 1973, in an article of Liu and Layland [3]. The Earliest Deadline First (EDF) algorithm is an analog of FCFS. Requests are ordered according to deadline and the request with the earliest deadline is serviced first. Assigning priorities to transactions an Earliest Deadline policy minimizes the number of late transactions in systems operating under low or moderate levels of resource and data contention. This is due to the highest priority given to the transactions that have the least remaining time in which to complete. However, the performance of Earliest Deadline steeply degrades in an overloaded system. Gaining high priority at this late stage may not leave sufficient time for transactions to complete before their deadlines. Under heavy loads, then, a fundamental weakness of the Earliest Deadline priority policy is that it assigns the highest priority to transactions that are close to missing their deadlines, thus delaying other transactions that might still be able to meet their deadlines. B. SCAN-EDF Algorithm It is the combination of two algorithms SCAN and EDF. When deadlines of two transactions are same then we applies this algorithm [4]. According to it, transactions with the same deadline are arranged using scan algorithms. Those who had shortest distance from disk head will get the An Improved Group-EDF: A Real-Time Disk Scheduling Algorithm S. Y. Amdani and M. S. Ali International Journal of Computer Theory and Engineering, Vol. 5, No. 6, December 2013 873 DOI: 10.7763/IJCTE.2013.V5.814