Investigating the Performance of Different Coding Schemes for Narrowband Internet of Things (NB-IoT) Applications

This study investigates and compares the effectiveness of five separate coding schemes for NB–IoT applications. The coding schemes investigated were Turbo, Low Density Parity Check (LDPC), Reed Solomon (RS), Turbo–RS hybrid code and LDPC–RS hybrid code. The performance of these schemes was assessed in both Additive White Gaussian Noise (AWGN) and Rayleigh fading channel conditions using Automatic Repeat Request (ARQ) method for error detection. LDPC codes employ a (1000 × 500) parity-check matrix and belief propagation for decoding, while turbo codes utilize pseudorandom data, block interleaving, and Viterbi decoding. Reed Solomon code employs generator polynomial in encoder and syndrome decoding. The five coding schemes were modelled and simulated using MATLAB software. The effectiveness of these codes was evaluated based on the Bit Error Rate (BER), Frame Error Rate (FER), and Energy Consumption. The simulation results show that all the coding schemes exhibit high energy consumption at low SNR for both channel conditions with RS consuming the most energy. In addition, the BER exhibits 10−1 value while the FER is near zero at low SNR for both channel conditions. At high SNR, the BER exhibits 10−6 value at 1.6 dB to 2.1 dB while the FER gives 10−1 value at 1 dB to 1.6 dB of SNR for both channels conditions with Turbo-RS hybrid code giving the best performance. This study indicates that Turbo-RS hybrid code is ideal for ultra-reliable, energy-efficient NB-IoT systems.