On channel estimation schemes for APD-based DDM-OFDM-PONs under sub-Nyquist sampling

Area of Science:

• Optical Communications
• Signal Processing
• Telecommunications Engineering

Background:

• Accurate channel estimation is crucial for delay-division multiplexing (DDM) orthogonal frequency-division multiplexing (OFDM) passive optical networks (PONs), especially under sub-Nyquist sampling constraints.
• Existing localized training schemes suffer from inaccurate estimations due to avalanche photodiode (APD) saturation variations.

Purpose of the Study:

• To propose and evaluate novel training schemes for channel estimation in APD-based OFDM-PONs operating under sub-Nyquist sampling.
• To improve the accuracy and efficiency of channel response estimation in DDM-OFDM-PON systems.

Main Methods:

• A distributed training scheme was proposed, distributing training subcarriers across frequency zones to mitigate APD saturation variations.
• An orthogonal training scheme was introduced, utilizing all subcarriers simultaneously with orthogonality between training symbols for efficient estimation.
• Experimental comparison of localized, distributed, and orthogonal schemes was conducted in a 25-Gbps APD-based OFDM-PON.

Main Results:

• The orthogonal scheme demonstrated superior performance in channel estimation accuracy compared to both localized and distributed schemes.
• The localized scheme exhibited the poorest channel estimation results.
• The orthogonal scheme enabled penalty-free operation in a DDM system at 1/32 Nyquist rate, achieving a 28 dB loss budget over 25 km fiber.

Conclusions:

• The orthogonal training scheme offers a significant improvement in channel estimation accuracy for DDM-OFDM-PONs under sub-Nyquist sampling.
• This approach effectively addresses the challenges posed by APD saturation variations, leading to enhanced system performance.
• The demonstrated penalty-free operation highlights the practical viability of the orthogonal scheme for high-speed optical networks.