This dissertation was presented to the Faculty of the Graduate School of The University of Texas at Austin in partial fulfillment of the requirements for the degree of

Ph.D. in Electrical Engineering


Graphical Models and Message Passing Receivers for Interference Limited Communication Systems  


Marcel Nassar, Ph.D.E.E.

The University of Texas at Austin, August 2013


Prof. Brian L. Evans


Dissertation - Defense Slides - Interference Mitigation Project - Smart Grid Communications Project


In many modern wireless and wireline communication networks, the interference power from other communication and non-communication devices is increasingly dominating the background noise power, leading to interference limited communication systems.

Conventional communication systems have been designed under the assumption that noise in the system can be modeled as additive white Gaussian noise (AWGN). While appropriate for thermal noise, the AWGN model does not always capture the interference statistics in modern communication systems. Interference from uncoordinated users and sources is particularly harmful to communication performance because it cannot be mitigated by current interference management techniques.

Based on previous statistical-physical models for uncoordinated wireless interference, this dissertation derives similar models for uncoordinated interference in PLC networks. The dissertation then extends these models for wireless and powerline interference to include temporal dependence among amplitude samples. The extensions are validated with measured data.

The rest of this dissertation utilizes the proposed models to design receivers in interference limited environments. Prior designs generally adopt suboptimal approaches and often ignore the problem of channel estimation which limits their applicability in practical systems. This dissertation uses the graphical model representation of the OFDM system to propose low-complexity message passing OFDM receivers that leverage recent results in soft-input soft-output decoding, approximate message passing, and sparse signal recovery for joint channel/interference estimation and data decoding. The resulting receivers provide huge improvements in communication performance (more than 10dB) over the conventional receivers at a comparable computational complexity. Finally, this dissertation addresses the design of robust receivers that can be deployed in rapidly varying environments where the interference statistics are constantly changing.


For more information contact: Marcel Nassar <>