TY - GEN
T1 - Performance evaluation of mobile wireless networks
T2 - ACM MSWiM 2001 Proceedings of the 4th ACM International Workshop on Modeling, Analysis and Simulation of Wireless and Mobile Systems
AU - Jobin, J.
AU - Faloutsos, Michalis
AU - Tripathi, Satish K.
PY - 2001
Y1 - 2001
N2 - We propose a methodology to simplify the analysis of wireless network simulations. Wireless simulation models are plagued by a vast parameter space. Consequently, performance studies are not easy to interpret or compare across different models. We reduce this parameter space by proposing a set of parameters that describe the network at a higher level of abstraction. More specifically, the cornerstone of our approach is the steady state utilization concept which quantifies the inherent capacity of a network to support a particular workload. The steady state utilization of a system is the maximum utilization with zero loss for a given traffic load. We also introduce the steady state arrival rate which is the arrival rate that keeps the network operating at its steady state utilization. Furthermore, we propose an efficient way to evaluate the performance of a network in terms of effective utilization which captures the utilization and the loss at the same time. Effective utilization measures the goodput of the network. Finally, as a case study, we use our methodology to evaluate how advance reservations affect network performance. We show how, contrary to intuition, bandwidth reservations can hurt performance, if we consider dropped calls as lost income.
AB - We propose a methodology to simplify the analysis of wireless network simulations. Wireless simulation models are plagued by a vast parameter space. Consequently, performance studies are not easy to interpret or compare across different models. We reduce this parameter space by proposing a set of parameters that describe the network at a higher level of abstraction. More specifically, the cornerstone of our approach is the steady state utilization concept which quantifies the inherent capacity of a network to support a particular workload. The steady state utilization of a system is the maximum utilization with zero loss for a given traffic load. We also introduce the steady state arrival rate which is the arrival rate that keeps the network operating at its steady state utilization. Furthermore, we propose an efficient way to evaluate the performance of a network in terms of effective utilization which captures the utilization and the loss at the same time. Effective utilization measures the goodput of the network. Finally, as a case study, we use our methodology to evaluate how advance reservations affect network performance. We show how, contrary to intuition, bandwidth reservations can hurt performance, if we consider dropped calls as lost income.
KW - Effective utilization
KW - Performance evaluation
KW - Reservation
KW - Steady state utilization
KW - Wireless network simulation
UR - https://www.scopus.com/pages/publications/0035789906
M3 - Conference contribution
SN - 1581133782
T3 - Proceedings of the 4th ACM International Workshop on Modeling, Analysis and Simulation of Wireless and Mobile Systems
SP - 65
EP - 72
BT - Proceedings of the 4th ACM International Workshop on Modeling, Analysis and Simulation of Wireless and Mobile Systems
A2 - Meo, M.
A2 - Dahlberg, T.A.
A2 - Donatiello, L.
Y2 - 21 July 2001 through 21 July 2001
ER -