CNN BASED RICIAN K FACTOR ESTIMATION FOR NON-STATIONARY INDUSTRIAL FADING CHANNEL

Abstract

Wireless networks attract increasing interests from a variety of industry communities. However, the wide applications of wireless industrial networks are still challenged by unreliable services due to severe multipath fading effects, especially the non-stationary temporal fading effect. Received Signal Strength Indicator (RSSI) will be a noisy estimation only on the specular power and fail to describe the link quality accurately without the aid of scattered power, while Rician K factor consisted by both the specular and scattered power can be treated as a reliable metric. The traditional estimation approaches of K factor from modulated wireless signals have to be data aided. In this paper, we attempt to formalize the estimation of K factor as a problem of non-linear feature extraction directly from modulated I/Q samples, which can be achieved through a simple convolutional neural network with morphological pre-processing. The experiments over field measurements have demonstrated the possibility of this methodology.