Wednesday, September 30, 2009

MACAW: A Media Access Protocol for Wireless LAN's

Summary

This paper presents a new protocol for wireless LAN's: MACAW. MACAW builds upon the MACA media access protocols by adding 2 new messages: Data-Sending and ACK resulting in a RTS-CTS-DS-DATA-ACK message exchange. MACAW also uses a different backoff algorithm.

The authors base their proposal on 4 observations of MACA proposal. 1) contention occurs at the receiver, not the sender 2) congestion is not homogenous, 3) all stations must work together to learn about the congestion levels and 4) the protocol should send info about contention periods to all devices so they can respond accordingly.

To make the use of bandwidth more fair, the authors propose a new backoff algorithm. In this algorithm, each of the stations shares their current value of the backoff counter so that all stations have the same backoff counter, resulting in fair allocation. Using multiple streams instead of single streams is another method to achieving fair allocation.

The authors add an ACK message which allows the receiver to tell the sender that they have received all the data. To solve the "exposed terminal" problem, they introduce the DS message which tells all nearby stations that a data transmission is about to occur. RRTS (Request-for-Request-to-Send) messages are also used to give all streams fair access to the media.

When evaluating MACAW, the authors found it to have an overhead of approximately 8% higher than MACA. However, it also has a throughput that is 37% higher than MACA, which more than makes up for the higher overhead.

Criticism & Questions

Although it wasn't highly critical to the protocol, I would be interested in knowing how they plan to change the packet header to accomodate the backoff counter value required for their backoff algorithm.

In addition, their backoff algorithm requires each station to copy over the backoff counter that they see in other stations' packets. I'm curious about how this could be misused. It doesn't seem like there's anyway to prevent a station from putting in a false value for other stations to use so that it can have an unfair advantage.

Feedback

I really enjoyed reading this paper. I think it was very well organized and easy to follow. I vote to keep it in the syllabus.

No comments:

Post a Comment