Performance analysis of AI aided embedded OFDM receiver based on RK3399 platform

Abstract

Orthogonal frequency division multiplexing (OFDM) technology is widely used to deal with the multipath effect. With the rise of artificial intelligence technology, deep learning method has been initially applied in the field of channel estimation and signal detection of OFDM system. This paper proposed an AI aided embedded OFDM receiver based on RK3399 platform. The channel estimation model based on CNN and BiLSTM is constructed, and PC platform and TB-RK 3399 embedded platform are used to train and deploy the proposed deep learning network model offline and online respectively. Experimental results show that, compared with the traditional LS and MMSE channel estimation methods, the proposed channel estimation method has a great improvement in estimation error and BER performance. The semi physical receiver based on TB-RK3399 achieves similar receiving performance with PC platform.