Finite-Blocklength Analysis of Coded Modulation with Retransmission.

Area of Science:

• Telecommunications Engineering
• Information Theory
• Wireless Communication Systems

Background:

• Rapid advancements in 5G and Beyond 5G (B5G) networks necessitate improved retransmission strategies.
• Existing finite-length coding performance prediction formulas require adaptation for coded modulation retransmission scenarios.
• High spectral efficiency retransmissions present unique challenges for performance prediction.

Purpose of the Study:

• To review and propose novel rate prediction formulas for coded modulation retransmission.
• To validate the accuracy of a recently proposed correction model in coded modulation retransmission contexts.
• To enhance the generality of prediction models by introducing a unified variable for varying SNRs.

Main Methods:

• Review of classical finite-length coding performance prediction formulas.
• Development of new rate prediction formulas tailored for coded modulation retransmission.
• Introduction of a range variable, Pfinal, to unify predictions across different signal-to-noise ratios (SNRs).
• Validation through extensive simulation results.

Main Results:

• The proposed rate prediction formulas demonstrate high accuracy for coded modulation retransmissions.
• The recently proposed correction model maintains high predictive accuracy in these advanced scenarios.
• The Pfinal variable effectively unifies prediction models across diverse SNR conditions.

Conclusions:

• The developed formulas and unified model provide accurate performance predictions for 5G/B5G retransmissions.
• Recommendations are provided for optimizing retransmission strategies, particularly for high spectral efficiency.
• The findings contribute to the efficient design and deployment of future wireless communication systems.