Month 1998 doc IEEE 802 11 99256 Evaluation

  • Slides: 9
Download presentation
Month 1998 doc. : IEEE 802. 11 -99/256 Evaluation of Protocol efficiency Alan Winkowski

Month 1998 doc. : IEEE 802. 11 -99/256 Evaluation of Protocol efficiency Alan Winkowski Breeze. COM Ltd. Submission 1 John Doe, His Company

Month 1998 doc. : IEEE 802. 11 -99/256 Introduction • New PHY’s defined lately

Month 1998 doc. : IEEE 802. 11 -99/256 Introduction • New PHY’s defined lately in 802. 11, while MAC remained static • Efficiency was analyzed for different packet sizes as well as different transaction scenarios • Analisys was performed on a transaction basis • Substantial degradation in protocol efficiency Submission 2 John Doe, His Company

November 1999 doc. : IEEE 802. 11 -99/256 Efficiency calculation • ND = Amount

November 1999 doc. : IEEE 802. 11 -99/256 Efficiency calculation • ND = Amount of Network Data • 60, 400 and 1500 bytes Data Packets • Tr. D = Transaction duration • DFC (Data-Ack and RTS-CTS-Data-Ack) • PCF (CF-Poll{+Data}-CF-Ack{+Data}) • NDR = Nominal Data Rate • 2 Mbps, 11 Mbps, 24 Mbps and 54 Mbps Submission 3 Alan Winkowski, Breeze. COM Ltd.

November 1999 doc. : IEEE 802. 11 -99/256 DFC Efficiency values • Events like

November 1999 doc. : IEEE 802. 11 -99/256 DFC Efficiency values • Events like Collisions and Backoff were not considered, thus efficiency is actually lower Submission 4 Alan Winkowski, Breeze. COM Ltd.

November 1999 doc. : IEEE 802. 11 -99/256 DCF: Data-ACK Submission 5 Alan Winkowski,

November 1999 doc. : IEEE 802. 11 -99/256 DCF: Data-ACK Submission 5 Alan Winkowski, Breeze. COM Ltd.

November 1999 doc. : IEEE 802. 11 -99/256 DFC: RTS-CTS-Data-Ack Submission 6 Alan Winkowski,

November 1999 doc. : IEEE 802. 11 -99/256 DFC: RTS-CTS-Data-Ack Submission 6 Alan Winkowski, Breeze. COM Ltd.

November 1999 doc. : IEEE 802. 11 -99/256 PCF Efficiency • Protocol performance depends

November 1999 doc. : IEEE 802. 11 -99/256 PCF Efficiency • Protocol performance depends on protocol flow • Flow simulation is required – packet length: 400 bytes – Mix of different transactions: • • Submission A. Data (CF-Poll) - Data (CF-Ack) B. CF-Poll - Data (with CF-Ack on next poll) C. Data (CF-Poll) - CF-Ack (No Data) D. CF-Poll (No Data) - PIFS 7 Alan Winkowski, Breeze. COM Ltd.

November 1999 doc. : IEEE 802. 11 -99/256 PCF Efficiency - simulation Submission 8

November 1999 doc. : IEEE 802. 11 -99/256 PCF Efficiency - simulation Submission 8 Alan Winkowski, Breeze. COM Ltd.

November 1999 doc. : IEEE 802. 11 -99/256 Conclusions and Recommendations • Quantitative efficiency

November 1999 doc. : IEEE 802. 11 -99/256 Conclusions and Recommendations • Quantitative efficiency degradation in all scenarios analyzed • Possible improvements proposed: – Greater frames for Data and Control information (specially in short frames) – Aggregation of feedback information – Scheduling in advance by the AP (PCF improvement) Submission 9 Alan Winkowski, Breeze. COM Ltd.