Computation Reduction for IRIS Process using DT-MSOF Method

Abstract

Defining the process of type IRIS (Increasing Reward with Increasing Service) is one effort to get around the limitations of time available for execution. In this paper, the Double-Track Most Significant Operation First (DT-MSOF) strategy is proposed to improve system efficiency. In this strategy, the order of execution of the computing element according to their importance and the execution need only be done until the result is correct and/or acceptable. To know the performance of this strategy, two simulations are performed and an example of application is demonstrated. The simulation results show that the execution efficiency using DT-MSOF is better than DT-LSOF and DT-RSOF, measured by the percentage of executed computation. The demonstration indicates the possibility of implementation of this strategy on particular problem.

DT-MSOF Strategy and its Application to Reduce the Number of Operations in AHP

“…One possible approach to shorten execution time is reducing the number of executed computation units. In [1] and [2] we proposed a new computation strategy to reduce the execution time, named Double Track -Most Significant Operation First (DT-MSOF). This strategy uses the concept of intermediate result, ordering executions based on their level of significance, uses an approximate computing approach, and restricts execution using an interval arithmetic concept.…”