Spectrally efficient digital mobile fronthaul with discrete cosine transform and multi-band quantization

Area of Science:

• Telecommunications Engineering
• Digital Signal Processing
• Optical Communications

Background:

• Digital mobile fronthaul (MFH) is crucial for 5G networks, but spectral efficiency remains a challenge.
• Existing digital MFH systems often require significant bandwidth for transmitting digitized symbols.
• Orthogonal frequency division multiplexing (OFDM) is a standard modulation technique in modern wireless systems.

Purpose of the Study:

• To propose and demonstrate a spectrally efficient digital MFH system.
• To reduce the number of quantization bits (QBs) required for digitizing OFDM symbols.
• To enhance the overall bandwidth efficiency of the mobile fronthaul link.

Main Methods:

• Utilizing discrete cosine transform (DCT) to transform baseband OFDM symbols into coefficients.
• Implementing multi-band quantization (MBQ) to assign different QBs to coefficients based on significance.
• Discarding less significant transform coefficients to further reduce data.
• Experimentally demonstrating the proposed DCT-MBQ digital MFH over a 25 Gb/s optical link.

Main Results:

• Achieved significant reduction in required QBs per OFDM symbol.
• Demonstrated high bandwidth efficiency through the DCT-MBQ approach.
• Successfully transmitted signals supporting up to 16x100 MHz 5G new radio carriers with 1024 QAM.
• Attained an error vector magnitude (EVM) of 0.65% in the experimental setup.
• Supported equivalent Common Public Radio Interface (CPRI) rates ranging from 55 to 158 Gb/s.

Conclusions:

• The proposed DCT-MBQ digital MFH offers a spectrally efficient solution for next-generation mobile networks.
• This method effectively reduces data requirements without compromising signal quality.
• The experimental validation confirms the viability and performance of the DCT-MBQ approach for high-capacity fronthaul.