Wednesday, August 31, 2016

Human identification using WiFi signal

Identification (and verification) in digital environments has become the Holy Grail of our time

It is relatively quick and easy to manage identities of users at local level. What complicates matters is deal with a structure able to identify users in decentralized environments in such way that we could optimize users experience and control their different permissions.      

And to make matters worse, the Internet of Things is forcing us to correspond digital word with the physical one so as not to be dependent on non standardized variables.

Under this rationales Ton Xin, Bin Guo, Zhu Wang, Mingyang Li and Zhiwen Yo, researchers from School of Computer Science, published a study which leverages WiFi signals to enable non-intrusive human identification in domestic environments. Experimental results indicate that the identification accuracy is about 88,9% to 94,5%, showing that the method is effective. 

The technique is based on the observation of each person, regarding their body shape characteristics and motion patterns. The influence is captured by the Channel State Information (CSI) and they use extra tools (Principal Component Analysis, Discrete Wavelet Transform and Dynamic Time Warping) that processes and analyses the signal and generate a pattern of behavior.

Its applications generate great opportunities in many sectors that are on everyone's lips today. After all, we are talking about a non-intrusive technology that can guess, quite probably, who is doing what, adjusting to the needs of each person.

In addition to this study regarding the indoor human identification with WiFi signals, in the last seven days we can also take into account the result of the research about the development and applications of RASP (Runtine Application Self-Protection) by Securisis experts and "Demystifying the Dark Web" by Danny Rogers from Terbium Labs.


Post a Comment