Title: Fuzzy Decision Analysis in Negotiation between the System of Systems Agent and the System Agent
Publication Date: 3/1/2013
Conference The International Journal of Software Engineering (JSCSE), Volume 3, NO. 3, Special Issue
Previous papers have described a computational approach to System of Systems (SoS) development using an Agent-Based Model (ABM). This paper describes the Fuzzy Decision Analysis used in the negotiation between the SoS agent and a System agent in the ABM of an Acknowledged SoS development. An Acknowledged SoS has by definition a limited influence on the development of the individual Systems. The individual Systems have their own priorities, pressures, and agenda which may or may not align with the goals of the SoS. The SoS has some funding and deadlines which can be used to negotiate with the individual System in order to illicit the required capability from that System.
The Fuzzy Decision Analysis determines how the SoS agent will adjust the funding and deadlines for each of the Systems in order to achieve the desired SoS architecture quality. The Fuzzy Decision Analysis has inputs of performance, funding, and deadlines as well as weights for each capability. The performance, funding, and deadlines are crisp values which are fuzzified.
The fuzzified values are then used with a Fuzzy Inference Engine to get the fuzzy outputs of funding adjustment and deadline adjustment which must then be defuzzified before being passed to the System agent. The first contribution of this paper is the fuzzy decision analysis that represents the negotiation between the SoS agent and the System agent. A second contribution of this paper is the method of implementing the fuzzy decision analysis which provides a generalized fuzzy decision analysis.