Performance evaluation of T‐COFDM under combined noise in PLC with log‐normal channel gain using exact derived noise distributions. Issue 6 (1st April 2019)

• Record Type: Journal Article
• Authors: Al‐Rubaye, Ghanim A.; Tsimenidis, Charalampos C.; Johnston, Martin
• Is Part Of: IET communications. Volume 13:Issue 6(2019)
• Page Start: 766
• Page End: 775
• Publication Date: 2019-04-01
• Subjects: Nakagami channels -- error statistics -- carrier transmission on power lines -- convolutional codes -- decoding -- impulse noise -- OFDM modulation -- Gaussian noise -- Monte Carlo methods -- turbo codes -- equalisers -- radiofrequency interference; performance evaluation -- log‐normal channel gain -- exact derived noise distributions -- performance analyses -- frequency‐selective power‐line communication -- complex‐valued ratio distributions -- effective noise samples -- Nakagami‐m background interference noise -- turbo convolutional code -- exact log‐likelihood ratio -- pairwise error probability -- bit error rate degradation -- conventional T‐COFDM system -- nonGaussian noise -- MCAIN model -- combined Nakagami‐m BI noise -- AUBs derivation -- BER improvements -- PLC -- turbo‐coded orthogonal frequency division multiplexing -- zero‐forcing equaliser output -- middleton class A impulsive noise -- soft decoder -- average upper‐bounds -- clipping thresholds -- Monte Carlo simulations -- analytical BER expression; Telecommunication systems -- Periodicals; Speech processing systems -- Periodicals; 621.38205
• DOI: 10.1049/iet-com.2018.6185
• Languages: English
• ISSNs: 1751-8628

Abstract

Abstract : In this study, the performance analyses of the proposed turbo‐coded orthogonal frequency division multiplexing (T‐COFDM) are investigated over the power‐line communication with lognormal channel gain based on derived effective complex‐valued ratio distributions of the individual and combined noise samples at the zero‐forcing equaliser output. The effective noise samples are derived in the presence of Nakagami‐m background interference (BI) noise, Middleton class A impulsive noise (MCAIN) and their combination. The performance of the turbo code has been improved by computing the exact log‐likelihood ratio using derived distributions, with the derivation of pairwise error probability and the average upper‐bounds (AUBs). Moreover, the bit error rate (BER) degradation in the conventional T‐COFDM system has been improved by deriving two clipping thresholds to combat the effect of the non‐Gaussian noise, the first one has been derived in the presence of the impulsive noise only modelled by MCAIN model and the second one in the presence of combined Nakagami‐m BI noise and MCAIN model. Monte Carlo simulations demonstrate the significant BER improvements of the proposed T‐COFDM system compared to the improved conventional T‐COFDM system with a close agreement to the AUBs derivation and analytical BER expression.