Performance comparison and complexity analysis of equalization techniques in MIMO NB-PLC

Communication over power lines has been an interesting subject in last two decades. PLC conducts information simultaneously on the same wire which is utilized for electric power transmission but within different frequency intervals, not 50 or 60 Hz which is reserved for AC power. Among the various benefits of this technique the most significant one is the considerable reduction in cost due to reusing the existing power network for communication means which is available almost everywhere. Multiple Input Multiple Output (MIMO) and OFDM techniques from the prospective of PLC and their role in improving the performance of the PLC are explained in this thesis. Since the power network was not initially designed for communicating purposes, we face harsh channel and noise during communication over this infrastructure. In addition, while using MIMO method for communication, there is interference present among the symbols. Thus, robust detection techniques are vital to reduce the noise effects and assist us in reconstructing the original signal. This will be done by equalization techniques. This thesis investigates an OFDM-based NB-PLC system. The work is based on IEEE 1901.2, which is "Standard for Low Frequency (less than 500 kHz) Narrow Band Power Line Communications for Smart Grid Applications". Pre-coded spatial multiplexing together with Post-processing techniques like Linear and Non-linear detection such as: Zero Forcing (ZF), Minimum Mean Square Error (MMSE), Successive interference cancellation (SIC), Ordered successive interference cancellation (OSIC), and Maximum likelihood (ML) are applied on 3 × 3 MIMO NB-PLC system in the AWGN (Additive White Gaussian Noise) channel and under different modulation schemes (BPSK , QPSK,8PSK, 16QAM) and the corresponding simulation results are discussed. In addition to evaluating equalizers’ performance according to the simulation results, we analyze the complexity of the algorithms in order to assess their applicability in real systems