Skip to Main Content
Modeling mobility and user behavior is of fundamental importance in testing the performance of protocols for wireless data networks. Although several models have been proposed in the literature, none of them can at the same time capture important features such as geographical mobility, user-generated traffic, and the wireless technology at hand. When collectively considered, these three aspects determine the user-perceived quality-of-service (QoS) level, which, in turn, might have an influence on the mobility of those users (we call them QoS-driven users) who do not display constrained mobility patterns, but they can decide to move to less congested areas of the network in case their perceived QoS level becomes unacceptable. In this paper, we introduce the Wireless QoS-aware Mobility (WiQoSM) model, which collectively considers all of the above mentioned aspects of wireless data networks. WiQoSM is composed of 1) a user mobility model, 2) a user traffic model, 3) a wireless technology model, and 4) a QoS model. Components 1,2, and 3 provide input to the QoS model, which, in turn, can influence the mobility behavior of QoS-driven users. WiQoSM is very simple to use and configure and can be used to generate user and traffic traces at the access points (APs) composing a wireless data network. WiQoSM is shown to be able to generate traces that resemble statistical features observed in traces extracted from real-world wireless local area network (WLAN) deployments. Furthermore, WiQoSM has the nice feature of allowing the fine tuning of a disjoint set of parameters in order to influence different statistical properties of the generated traces and of providing the network designer with a high degree of flexibility in choosing network parameters such as the number of users and APs, wireless channel technology, traffic mix, and so on. Given the above features, WiQoSM can be a valuable tool in the simulation of wireless data network protocols.
Date of Publication: Feb. 2008