Advanced Signal Processing Techniques for Smart Grid Communications
Smart grid (SG) networks require a reliable communication infrastructure for bi-directional communication. The well-established network infrastructure which covers a vast geographical area motivates SG communications. The SG is expected to incorporate a hybrid mesh of different communication technologies over different physical media to create an efficient and reliable communication network. This dissertation focuses on the unlicensed wireless communication and power-line communication (PLC) approaches to SG communications.
Wireless SG smart utility networks (SUNs) use unlicensed frequency bands which motivate the need for wireless coexistence mechanisms and reliable spectrum sensing. We investigate the spectrum sensing problem in orthogonal frequency-division multiplexing (OFDM) based cognitive radio networks under In-phase/Quadrature (I/Q) imbalance. We derive the likelihood ratio test (LRT) in the presence of I/Q imbalance at the analog front-ends of both the primary and secondary users. In addition, we derive closed-form expressions for the probabilities of detection and false alarm and the receiver operating characteristics and examine their dependence on both transmit and receive I/Q imbalance levels. Furthermore, we compare the performance of the LRT with that of the energy detector and demonstrate the superiority of the former over the latter. Next, we generalize our analysis to the blind case where we derive simple closed-form expressions for the generalized likelihood ratio test (GLRT) and its false alarm probability as a function of the received signal only, i.e., without requiring any knowledge of the primary-to-secondary channel response, noise statistics, or I/Q imbalance parameters. Our results demonstrate that the correlation properties of the primary user's signal induced by transmit I/Q imbalance are signal features that can be exploited in a blind fashion at the secondary user to enhance the detection probability significantly compared to the conventional energy detector.
Furthermore, we propose and analyze a multiple-input multiple-output (MIMO) OFDM-based bi-directional communication system for SG applications in medium-voltage (MV) PLC. Multi-conductor transmission line theory is used for accurate narrowband (NB) MV-PLC channel characterization and evaluation of the spatial correlation between the phase conductors. The proposed MV MIMO-OFDM design can achieve significant data rate gains over single-input single-output OFDM transmission by optimizing the indices of the active OFDM sub-channels, the bit allocation across them, and the transmit power allocation across the spatial beams. The achievable MIMO-OFDM data rates are evaluated for an MV overhead network with different line lengths, coupler impedances, and branches. In addition, we thoroughly investigate the possibility of reliable data transmission through distribution transformers, utilizing a MIMO NB-PLC OFDM-based approach. The achievable data rates are quantified for NB MV-PLC line in cascade connection with distribution transformers.
We investigate the multi-user sum-rate optimization for point-to-multipoint Multicasting MV PLC network. We design the optimal spatial precoder and equalizer which maximize the achievable sum-rate for all users in the network. In addition, we propose two schemes to secure data transmission through the two-group multicasting network. The first scheme secures the network when the eavesdropper is a user in one of the two multicasting groups. The second scheme secures the multicasting network from external passive eavesdroppers. The secrecy rate is evaluated for both scenarios.