Congestion avoidance and control
SIGCOMM '88 Symposium proceedings on Communications architectures and protocols
SIGMETRICS '97 Proceedings of the 1997 ACM SIGMETRICS international conference on Measurement and modeling of computer systems
On the generation and use of TCP acknowledgments
ACM SIGCOMM Computer Communication Review
A capacity analysis for the IEEE 802.11 MAC protocol
Wireless Networks
Insensitivity in processor-sharing networks
Performance Evaluation
Performance of a shared packet wireless network with interactive data users
Mobile Networks and Applications
Insensitive Bandwidth Sharing in Data Networks
Queueing Systems: Theory and Applications
Performance analysis of the IEEE 802.11 distributed coordination function
IEEE Journal on Selected Areas in Communications
Performance modelling and measurements of TCP transfer throughput in 802.11-based WLAN
Proceedings of the 9th ACM international symposium on Modeling analysis and simulation of wireless and mobile systems
Computer Networks: The International Journal of Computer and Telecommunications Networking
Characterizing the exit process of a non-saturated IEEE 802.11 wireless network
Proceedings of the tenth ACM international symposium on Mobile ad hoc networking and computing
Average-value analysis of 802.11 WLANs with persistent TCP flows
IEEE Communications Letters
Hi-index | 0.00 |
We consider an IEEE 802.11 based wireless LAN where an access point is used to connect a fixed number of users to the Web or to a shared file system. Users alternate between activity periods (corresponding to the download of a file) and idle periods (corresponding to think times). We first consider the interaction of TCP with the IEEE 802.11 MAC protocol, and get approximate expressions for the TCP throughput in the presence of nc competing persistent connections. This result is then used to develop a queueing model, which provides the mean session delay in the presence of short-lived flows. The analysis also accounts for the TCP delayed ACK option. Comparison with simulation outcomes is provided, validating the model and providing guidelines for network designers. A particular emphasis is devoted to the impact of the TCP's advertised window size; in particular, it is proved that setting it to a small value leads to insensitivity of mean file transfer times to the file size distribution.