AI-Aided Online Adaptive OFDM Receiver: Design and Experimental Results

TL;DR

From the OTA test, the AI-aided OFDM receivers, especially the SwitchNet receiver, are robust to OTA environments and promising for future communication systems.

Abstract

Orthogonal frequency division multiplexing (OFDM) has been widely applied in current communication systems. The artificial intelligence (AI)-aided OFDM receivers are currently brought to the forefront to replace and improve the traditional OFDM receivers. In this study, we first compare two AI-aided OFDM receivers, namely, data-driven fully connected deep neural network and model-driven ComNet, through extensive simulation and real-time video transmission using a 5G rapid prototyping system for an over-the-air (OTA) test. We find a performance gap between the simulation and the OTA test caused by the discrepancy between the channel model for offline training and the real environment. We develop a novel online training system, which is called SwitchNet receiver, to address this issue. This receiver has a flexible and extendable architecture and can adapt to real channels by training only several parameters online. From the OTA test, the AI-aided OFDM receivers, especially the SwitchNet receiver, are robust to real environments and promising for future communication systems. We discuss potential challenges and future research inspired by our initial study in this paper.

Figures (12)

• Figure 1: Block diagram of an OFDM system, where pilot symbols are inserted at the transmitter, and the receiver acquires CSI. Compared with the traditional OFDM receiver, the AI receiver performs CE, SD, and QAM demodulation altogether and directly maps the received signals into decided symbols.
• Figure 2: Structure of FC-DNN. The FC-DNN receiver contains five FC layers that directly map the received signal to the recovered bitstreams.
• Figure 3: ComNet receiver architecture. The two subnets use traditional communication solutions as initializations and apply DL networks to refine the coarse inputs. The crossing short-path provides a relatively robust candidate of the binary symbol recovery.
• Figure 4: CE subnet architecture of the SwitchNet receiver. The CE RefineNet 0 is the basic DNN network for CE, and the CE RefineNets from 1 to $M$ are the compensating network of the CE RefineNet 0. $\alpha$ is the switch parameters to decide whether the CE RefineNets from 0 to $M$ are accessed.
• Figure 5: Frame structure of the simulated OFDM system. Each frame contains several OFDM symbols. A pilot symbol and a data symbol are set as the inputs of NNs. Each OFDM symbol contains 128 subcarriers. A total of 64 subcarriers are used for the pilot symbol or data symbol transmission while the vest 64 subcarriers serve as guard band and DC offset.