Revisit of RTS/CTS exchange in high-speed IEEE 802.11 networks

TLDR: It is concluded that, for the heterogeneous data rate environments, the RTS/CTS threshold should be redefined as a frame transmission time rather than as a frames size.

Abstract: IEEE 802.11 medium access control (MAC), called distributed coordination function (DCF), provides two different access modes, namely, 2-way (basic access) and 4-way (RTS/CTS) handshaking. The 4-way handshaking has been introduced in order to combat the hidden terminal phenomenon. It has been also proved that such a mechanism can be beneficial even in the absence of hidden terminals, because of the collision time reduction. We analyze the effectiveness of the RTS/CTS access mode, in current 802.11b and 802.11a networks. Since the rates employed for control frame transmissions can be much lower than the rate employed for data frames, the assumption on the basis of the 4-way handshaking introduction, i.e., a short transmission time for the RTS control frame, is no longer valid. As a consequence, the basic access mode results in the optimal access solution in most cases, even in heavy load conditions with hidden nodes. We compare the 2-way and 4-way access performances through both analytical and simulation tools. We also discuss the operating conditions at which the switch from one access mode to another is desired for the cases of uniform and heterogeneous data rates among the stations. We conclude that, for the heterogeneous data rate environments, the RTS/CTS threshold should be redefined as a frame transmission time rather than as a frame size.