Today's economy creates the need for dynamic, adaptive and autonomous building of enterprise value chains consisting of arbitrary virtualized computing resources, as hardware and software services. The current key technology for service provisioning is the cloud computing framework. In the course of this development digital service markets are becoming business reality. Consumers in these digital markets apply typically the classical "take-it-or-leave-it" supermarket approach, which limits market performance. A solution to this problem is the so called Bazaar-based market, which extends the classical supermarket approach by enabling multi-round negotiation processes. Hereby business strategies of providers and consumers can be reflected in the negotiation processes to allow for smarter and more effective agreements in the market. In this paper we present a novel genetic algorithm based multi-round negotiation strategy between providers and consumers of services. This approach is realized within a Bazaar-Extension for CloudSim. To compare the market efficiency of resource allocations we introduce a novel key figure, the Bazaar-Score metric, which allows the evaluation of different business strategies.

Top