April 15, 2015

Speaker: Visal Chea, UOIT Computer Science graduate student

Title: Hamming Distance as a Metric for the Detection of Side Channel in 802.11 Wireless Communications

Abstract: Wireless technology has become a main player in communication through its desirable characteristics of mobility. Like many technologies, there are ways it can be exploited. One of these ways is through side channel communication where secret messages are passed along by purposely corrupting frames. These side channels can be established through intentionally corrupting the Frame Check Sequence (FCS) field by using a Cyclic Redundancy Check (CRC) polynomial which is different from standard. Malicious nodes exploit the fact that normal unsuspecting nodes will immediately drop these frames. In other words, if the frame is transmitted with error, it is deemed corrupted and dropped immediately by the receiver without any further inspection. In order to detect these side channels, there has to be a metric or feature that can be used to distinguish legitimate from illegitimate errors. The proposed feature in this thesis for detection of this type of side channel is the Hamming distance metric. The Hamming distance measures the number of bit differences between two bit strings. The proposed detection method is to apply this Hamming distance (HD) measure to compare CRC values that are generated by different CRC polynomials. The hypothesis is that the average HD between two CRC values generated by two different CRC polynomial would be significantly far apart than the average HD of CRC value of a frame that was naturally corrupted but was generated by the same CRC polynomial. In order to test this hypothesis the Hamming distance metric was first validated by using it as a feature in the Perceptron and Pocket algorithm. It was then evaluated for its effectiveness of detection using a well‐known score called the F‐Score. In order to obtain the data for validation and evaluation of the Hamming distance metric, a hybrid testing approach was introduced, which combined real data capture with simulation. The final results show very good promise that this Hamming distance metric is effective in detecting the presence of side channel communication.