Use this URL to cite or link to this record in EThOS:
Title: Coded-OFDM for PLC systems in non-Gaussian noise channels
Author: Al-Rubaye, Ghanim Abdulkareem Mughir
ISNI:       0000 0004 7429 9404
Awarding Body: Newcastle University
Current Institution: University of Newcastle upon Tyne
Date of Award: 2017
Availability of Full Text:
Access from EThOS:
Access from Institution:
Nowadays, power line communication (PLC) is a technology that uses the power line grid for communication purposes along with transmitting electrical energy, for providing broadband services to homes and offices such as high-speed data, audio, video and multimedia applications. The advantages of this technology are to eliminate the need for new wiring and AC outlet plugs by using an existing infrastructure, ease of installation and reduction of the network deployment cost. However, the power line grid is originally designed for the transmission of the electric power at low frequencies; i.e. 50/60 Hz. Therefore, the PLC channel appears as a harsh medium for low-power high-frequency communication signals. The development of PLC systems for providing high-speed communication needs precise knowledge of the channel characteristics such as the attenuation, non-Gaussian noise and selective fading. Non-Gaussian noise in PLC channels can classify into Nakagami-m background interference (BI) noise and asynchronous impulsive noise (IN) modelled by a Bernoulli-Gaussian mixture (BGM) model or Middleton class A (MCA) model. Besides the effects of the multipath PLC channel, asynchronous impulsive noise is the main reason causing performance degradation in PLC channels. Binary/non-binary low-density parity check B/NB-(LDPC) codes and turbo codes (TC) with soft iterative decoders have been proposed for Orthogonal Frequency Division Multiplexing (OFDM) system to improve the bit error rate (BER) performance degradation by exploiting frequency diversity. The performances are investigated utilizing high-order quadrature amplitude modulation (QAM) in the presence of non-Gaussian noise over multipath broadband power-line communication (BBPLC) channels. OFDM usually spreads the effect of IN over multiple sub-carriers after discrete Fourier transform (DFT) operation at the receiver, hence, it requires only a simple single-tap zero forcing (ZF) equalizer at the receiver. The thesis focuses on improving the performance of iterative decoders by deriving the effective, complex-valued, ratio distributions of the noise samples at the zeroforcing (ZF) equalizer output considering the frequency-selective multipath PLCs, background interference noise and impulsive noise, and utilizing the outcome for computing the apriori log likelihood ratios (LLRs) required for soft decoding algorithms. On the other hand, Physical-Layer Network Coding (PLNC) is introduced to help the PLC system to extend the range of operation for exchanging information between two users (devices) using an intermediate relay (hub) node in two-time slots in the presence of non-Gaussian noise over multipath PLC channels. A novel detection scheme is proposed to transform the transmit signal constellation based on the frequency-domain channel coefficients to optimize detection at the relay node with newly derived noise PDF at the relay and end nodes. Additionally, conditions for optimum detection utilizing a high-order constellation are derived. The closedform expressions of the BER and average BER upper-bound (AUB) are derived for a point-to-point system, and for a PLNC system at the end node to relay, relay to end node and at the end-to-end nodes. Moreover, the convergence behaviour of iterative decoders is evaluated using EXtrinsic Information Transfer (EXIT) chart analysis and upper bound analyses. Furthermore, an optimization of the threshold determination for clipping and blanking impulsive noise mitigation methods are derived. The proposed systems are compared in performance using simulation in MATLAB and analytical methods.
Supervisor: Not available Sponsor: Ministry of Higher Education in Iraq
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available