Research Interests
My research focuses on channel coding for communication systems. Channel coding is the art and science of encoding data such that:
it is less likely to become corrupted during the communication process. This is accomplished by encoding the data such that it has characteristics that closely match the time- or frequency-domain constraints of the communication channel. Coding techniques designed for this purpose are called line codes (in transmission systems), recording codes (in magnetic and optical recording systems), modulation codes, or more generally, constrained sequence coding techniques.
in the event that data becomes corrupted during the communication process, this corruption can be detected at the receiver, and measures can be taken to improve the accuracy of the received symbol sequence. These techniques are collectively known as error control coding techniques.
Ongoing research projects include:
further development of multimode line coding techniques, including guided scrambling. Multimode codes are of significant interest in applications where highly efficient line codes are required and/or where low complexity encoding and decoding processes are of interest. They have also been shown to be very useful in generating deep spectral nulls at DC through their configuration as high-order spectral-null codes.
integration of line coding and error control coding techniques. Standard concatenation of error control and line coding procedures suffers from the fact that error propagation that occurs during the line decoding operation impacts the performance of the error control decoder. Our ongoing efforts integrate and/or reverse the order of these decoding operations such that the detrimental impact of this error propagation is avoided.
development of efficient architectures for turbo decoding and other iterative error control decoding techniques. Powerful error control codes that can be decoded through iterative techniques have been proven to provide performance that approaches the Shannon limit. We are investigating novel architectures for the efficient implementation of these decoding techniques.
design of low density parity check codes for multiple-input multiple-output wireless applications. Most LDPC codes have been developed based on the assumptions of binary signaling over the Gaussian channel. We are considering the design of LDPC codes for higher-order alphabets, given the characteristics of the MIMO wireless channel.
application of constrained coding techniques to reduction of the peak-to-average power ratio in orthogonal frequency division multiplexed wireless systems. OFDM exhibits significant potential in future wireless applications, however its high inherent PAPR has limited its adoption. It has been shown that it is possible to bound the PAPR by encoding the data prior to the constructing the time-domain signal. By interpreting these encoding techniques as constrained sequence codes, we are applying knowledge from the constrained coding community to this wireless application.