Adaptive space-time compression for efficient massive MIMO fronthauling

Area of Science:

• Telecommunications Engineering
• Signal Processing
• Optical Communications

Background:

• Fifth-generation (5G) networks require massive MIMO antennas for high traffic density.
• Conventional fronthaul interfaces (e.g., CPRI) lead to linearly increasing bandwidth demands with antenna count.
• Optical fronthaul links face bandwidth limitations due to cost-effective components and physical constraints like chromatic dispersion.

Purpose of the Study:

• To propose and validate an adaptive space-time compression technique for improving fronthaul bandwidth efficiency in C-RAN.
• To decouple fronthaul bandwidth requirements from the number of antennas, making them dependent on user count.
• To investigate the joint optimization of the proposed compressor with optical fronthaul links.

Main Methods:

• Development of an adaptive spatial filter to reduce spatial channels by tracking the signal subspace.
• Implementation of adaptive quantizers for time-domain bandwidth compression of each channel.
• Experimental validation of the combined space-time compressor and optical fronthaul link.

Main Results:

• The proposed technique makes fronthaul bandwidth dependent on the number of users, not antennas.
• Experimental results demonstrated support for 32 users with OFDM signals over a 10GBd PAM4 optical interface.
• Achieved a 259.5-Gb/s CPRI-equivalent rate with less than 1% EVM for uplink 256-antenna fronthaul.

Conclusions:

• The adaptive space-time compression technique significantly enhances fronthaul bandwidth efficiency for 5G C-RAN.
• Joint optimization with optical fronthaul enables cost-effective high-capacity mobile communication.
• The method offers increased flexibility and feasibility for joint space-time optimization compared to traditional methods.