Trust Based Framework for Information Sharing Behavior in Command and Control Environments.

摘要

Information in tactical networks is being generated at an ever increasing pace. This is requiring networks to distribute the information processing in an efficient and intelligent way. Command and control (C2) theory suggests that the performance of such networks or organizations is dependent on the overall topology of the network and other parameters or characteristics of the constituent nodes, such as their ability filter information. Traditional information sharing policies use static approaches and operate on a need-to-know basis. The more network-enabled approach is to share-before-process, which introduces more dynamism and robustness to the system. However, no model exists to this day to appropriately study the trade-offs introduced by these different organizational structures.This paper proposes the use of an agent model in which agents consider two constructs, willingness and competence, when deciding with whom to share information. We show that these two constructs are general enough to capture organizational structures based on strict as well as shared governance rules. Furthermore, these constructs have clear analogues in cognitionpsychology, which allows us to incorporate trust into the agentmodels in a seamless manner. Through simulation, we study an information sharing scenario based on this agent model in different organizational structures. We illustrate the trade-offs between situation awareness (SA) gained and network overhead in terms of communications and time required to reach a steady state SA.