Classifying Rate Adaptation Algorithms in IEEE 802.11b/g/n Wireless Networks

This thesis provides a tool for helping wireless network professionals to recognize different Rate Adaptation Algorithms (RAA) implemented in IEEE 802.11 wireless network device drivers. The RAA used in the wireless network adapters are responsible for selecting the bit-rate used by the hardware when transmitting frames over the wireless channel.This algorithm heavily affects the performance of wireless devices. We present the Rate Adaptation Classifier (RAC) which passively listens to data traffic between wireless stations. Based on the observed traffic, RAC performs logging and statistics. The final output of the application can be used to classify the rate adaptation algorithm used by the observed wireless device. RAC can be used on any platform which exports the correct headers to user-space through the PCAP framework. RAC has the ability to listen to and analyse any IEEE 802.11b/g wireless network and contains code to perform basic statistics on IEEE 802.11n. RAC captures and logs the important pieces of observed data traffic and is not affected by the encryption used by the wireless network. RAC is only interested in the Physical (PHY) and some Link-Layer information exported by the monitor interface. We present a series of validation tests, analyse the results obtained from RAC and compare these results against the theoretical expected behaviour of each RAA. We show that the results produced are directly comparable to the RAAs expected behaviour. We will also present the results from a series of experiments where we test Rate Adaptation Classifier (RAC) and the proposed method to match its output to known rate adaptation algorithms