Adaptive Audio Error Concealment Leveraging BER and PSNR Variability

In the sequential recovery framework, missing audio samples from transmitted G.723.1 codec data are sequentially reconstructed using adaptive Error Concealment (EC) techniques. The flexibility of this adaptive EC predictor enables the recovery of significant audio features, ensuring superior quality reconstruction. The proposed method consistently achieves Peak Signal-to-Noise Ratio (PSNR) values as high as 38.11 dB at low Bit Error Rate (BER) levels, whereas existing techniques exhibit significant variability, with PSNR values dropping to 22.47 dB at higher BER levels. This paper explores various EC approaches applied in audio codecs, analyzing their strengths, limitations, and associated research gaps. Finally, the study provides a comprehensive review and analysis of current approaches, highlights critical research challenges, and suggests actionable solutions to advance the state of resilient audio transmission systems.