TY - GEN
T1 - Self similarity analysis and modeling of VoIP traffic underwireless heterogeneous network environment
AU - Canberk, Berk
AU - Oktug, Sema
PY - 2009
Y1 - 2009
N2 - Self similar behavior of the aggregate traffic is a wellknown issue in the networking area. In this paper, we first study the self similarity of the empirically aggregated VoIP traffic in a heterogeneous wireless network testbed environment and then model it using Fractional Gaussian Noise (fGn). The heterogeneity of the environment is provided by exploiting different wireless technologies in backbone and access networks. The backbone of the testbed is IEEE 802.16d WiMAX whereas the access network is IEEE 802.11b WiFi mesh architecture. We evaluate the self similarity in terms of throughput and packet inter-arrival time using empirically captured VoIP calls generated by softphones in the laboratory. We prove the collected data's self similar characteristics with stochastic analysis using autocorrelation functions. We implement three well-known timedomain estimators to obtain Hurst values for both metrics. We also suggest the Fractional Gaussian Noise (fGn) Model for the empirically aggregated VoIP data. The self similarity analysis and modeling performed in this work will motivate new design issues on the quality of service frameworks and resources allocation mechanisms such as buffers in wireless heterogeneous networks.
AB - Self similar behavior of the aggregate traffic is a wellknown issue in the networking area. In this paper, we first study the self similarity of the empirically aggregated VoIP traffic in a heterogeneous wireless network testbed environment and then model it using Fractional Gaussian Noise (fGn). The heterogeneity of the environment is provided by exploiting different wireless technologies in backbone and access networks. The backbone of the testbed is IEEE 802.16d WiMAX whereas the access network is IEEE 802.11b WiFi mesh architecture. We evaluate the self similarity in terms of throughput and packet inter-arrival time using empirically captured VoIP calls generated by softphones in the laboratory. We prove the collected data's self similar characteristics with stochastic analysis using autocorrelation functions. We implement three well-known timedomain estimators to obtain Hurst values for both metrics. We also suggest the Fractional Gaussian Noise (fGn) Model for the empirically aggregated VoIP data. The self similarity analysis and modeling performed in this work will motivate new design issues on the quality of service frameworks and resources allocation mechanisms such as buffers in wireless heterogeneous networks.
UR - http://www.scopus.com/inward/record.url?scp=70349731435&partnerID=8YFLogxK
U2 - 10.1109/AICT.2009.19
DO - 10.1109/AICT.2009.19
M3 - Conference contribution
AN - SCOPUS:70349731435
SN - 9780769536118
T3 - Proceedings of the 2009 5th Advanced International Conference on Telecommunications, AICT 2009
SP - 76
EP - 82
BT - Proceedings of the 2009 5th Advanced International Conference on Telecommunications, AICT 2009
T2 - 2009 5th Advanced International Conference on Telecommunications, AICT 2009
Y2 - 24 May 2009 through 28 May 2009
ER -