A decision support system for distributed data allocation in support of C/sup 3/ systems

The allocation of data in distributed database system can significantly affect the response time of the system. This is particularly important when the database supports an operational command, control, and communication C/sup 3/ system. There are several methods for determining where to store the data in a distributed system. These include various optimization algorithms which use linear programming to minimize some cost function, as well as numerous heuristic methods. A decision support system (DSS) is presented that rank-orders various allocation methods based on the relative importance given to such factors as response times in the communication network, resources needed, and execution time of the allocation algorithm. To demonstrate this DSS the US army maneuver control system (MCS) was simulated using SLAM. Two optimal allocation schemes (using linear programming) and three heuristic methods were implemented to assign data location among the MCS nodes. The results of these simulations were then input to the DSS, along with several scenarios to provide importance weights to the decision factors. The results show that these weights affect the choice of allocation scheme, including some scenarios where neither of the optimal techniques were selected.<
>