Abstract—In the same way that a biological fingerprint operates, radiometric fingerprinting can be used to designate the unique transmitter of a given signal to improve the security and privacy of wireless communication, using only external feature measurements to match with the feature library. Fingerprinting reflects the unintentional modulation of radio, and it contains rich non-ideal characteristics of internal components within emitter. Our previous research work on finding out effective fingerprintings to establish the unique identity of a transmitter. However, in practical use we find that the change of intentional modulation parameters will affect the working status of internal components, to further affect radiometric fingerprinting. This paper uses information dimension and box dimension of fractal dimension theory to extract the signals fingerprinting to identify three different radios. A signal acquisition system is designed to capture the signals to analyze the effect of modulation parameters on the fingerprinting, where signals are generated under three different modulation modes, signal bandwidths and transmission frequencies. The experimental results show that, firstly, the distribution of radio's fractal dimension feature is different under different modulation modes, secondly, signal bandwidth has little influence on the feature, thirdly, the change of transmission frequency will result in a great degree of drift on the feature.

 

Index Terms—radiometric fingerprinting, modulation parameters, fractal, box dimension, information dimension

Cite: Shouyun Deng, Zhitao Huang, Guodong Yang, Xueqiong Li, and Gangson Ding, "Effect of Modulation Parameters on Radiometric Fingerprinting," International Journal of Signal Processing Systems, Vol. 5, No. 1, pp. 18-23, March 2017. doi: 10.18178/ijsps.5.1.18-23