Joint Optimization of Digital Fronthaul Based on Delta-Sigma Modulation and Equal Length Level Conversion

In conventional mobile fronthaul (MFH) links, the radio-over-fiber (RoF) interface and optical transmission are separately optimized. In this work, we propose an equal-length level conversion (ELLC) scheme to jointly optimize the delta-sigma modulation (DSM) quantization noise and optical transmission performance. The ELLC scheme converts the high-level quantizer output of the DSM into a low-level signal with equivalent entropy using entropy coding. Forward error correction (FEC) is incorporated to guarantee successful decoding without reducing the data rate. Under the peak-power constraint, both the signal-to-noise ratio (SNR) of the recovered wireless signal and the received optical power (ROP) sensitivity of the optical fronthaul transmission are improved with the proposed ELLC scheme. We experimentally demonstrate the conversion of a 6-level quantized DSM waveform into a 4-level signal (6-to-4 scheme), and the conversion of an 8-level quantized DSM signal into a 6-level signal (8-to-6 scheme) at an over-sampling rate (OSR) of 8 in an optical intensity modulation-direct detection (IM-DD) link. Compared to the 4-level quantized DSM, the 6-to-4 scheme provides an 11-dB SNR improvement for wireless signals and a 3-dB enhancement in ROP sensitivity after 10-km standard single-mode fiber (SSMF) transmission in the O-band. Similarly, the 8-to-6 scheme achieves a 4-dB SNR improvement and a 2.5-dB enhancement in ROP sensitivity compared to the 6-level quantized DSM. Moreover, we demonstrate high-pass (HP) DSM based on the proposed ELLC scheme, revealing its advantage with flexible passband selection. By converting higher-order PAM to lower-order PAM, the ELLC scheme achieves both the high recovered wireless signal SNR associated with high quantization levels and the enhanced noise resilience of lower modulation formats in optical channels. The proposed ELLC approach is anticipated to be a promising candidate for digital MFH in the 5G-Advanced era, especially for cost-sensitive and high-fidelity applications.