On the impact of employing various network architectures on the performance of collaborative applications

Wireless networks are rapidly emerging to be the network architecture of choice due to their adaptability and scalability. However, the lack of acceptable Quality of Service (QoS) properties has prevented the widespread use of wireless networks in several time-critical applications. Hence, it has become increasingly important to understand the relative performance levels of different types of networks on real world applications. This paper aims at testing the suitability and the impact of different network modes and architectures on the performance of collaborative Group Decision Support Systems (GDSS) applications. Collaborative GDSS applications support decision-making by connecting group of decision makers through computer network. These types of applications pose heavy demands on networks requiring minimum response time, minimum latency and maximum throughput for group collaboration. We propose a hybrid simulation approach used for the evaluation of key performance statistics of real world collaborative applications without the need for extensive resource investment spend on full application deployment in the target network. We implemented two network modes in three different network architectures by varying application loads. The obtained results shed the light on the impact of selecting different network architectures on the performance of GDSS applications.