Optimization of Data Link Layer Protocol Performance in High-EMI Communication Channels

The performance of data link layer (OSI Layer 2) protocols is critically affected by channel noise, electromagnetic interference (EMI), and propagation delay across diverse communication environments. This paper presents a rigorous mathematical framework for modeling and optimizing the throughput of Automatic Repeat reQuest (ARQ) protocols, namely Stop-and-Wait (SAW), Go-Back-N (GBN), and Selective Repeat (SR), operating over noisy channels. Analytical closed-form expressions for throughput efficiency η and effective goodput V are derived as functions of the Bit Error Rate (BER), frame size N, channel overhead C, transmission rate R, and normalized propagation delay. We further develop an exact optimization condition and an approximate closed-form solution for the optimal frame size Nopt ≈ √C/pb that maximizes system goodput. The integration of Cyclic Redundancy Check (CRC) for error detection and convolutional coding with Viterbi decoding for Forward Error Correction (FEC) is incorporated into the performance model, yielding a unified cross-layer analysis. Numerical results are validated against simulation data and applied to high-EMI railway communication standards (GSM-R, TETRA, FRMCS), alongside constrained IoT networks (LPWAN) and ultra-high-speed Data Center Networks (DCNs). The results demonstrate that adaptive frame sizing and robust error control can increase goodput by up to 47% in high-EMI environments compared to fixed-size frame strategies. Index Terms—ARQ protocols, data center networks, data link layer, error control coding, frame size optimization, good-put, IoT, mathematical modeling, noisy channels, railway communications.

Перевод пока недоступен