Comparative Study of Fixed-Rate Coded and Rateless Coded Cooperative Wireless Networks

Fixed-rate coding for cooperative wireless networks with decode-and-forward (DF) relay processing has received much attention. Yet, only a fairly narrow body of research exists on rateless coded DF relaying networks. Little is known about how the performances and complexity implications of fixed-rate and rateless coded systems compare. Meanwhile, such comparisons play a vital role in the system design for cooperative networks. This project aims at casting light on the relative merits of fixed-rate and rateless coded relaying networks in terms of data throughput and complexity, by analysis and simulation. Two matching, state-of-the-art fixed-rate and rateless coded relaying networks will be analyzed, simulated, and compared in terms of achievable data rates, the amount of CSI and feedback required at different nodes, and sensitivity to CSI estimation errors. Low density parity check (LDPC) codes, or a similar match, for the fixed-rate coded scheme and Raptor codes, or a variation, for the rateless coded scheme will be considered. The project commences with studying single-relay networks. Depending on the progress of the analytical and simulation parts of the project, multi-relay networks and networks in which relays are also sources of information will be examined.

The student will be primarily responsible for a comprehensive literature survey and constructing a comprehensive bibliography of the work that has been performed to date. The student will also be responsible for the computer simulation of the fixed-rate and rateless coded schemes, to determine their error probabilities, maximum data throughputs, and the sensitivity of the data throughputs to channel fading and CSI estimation errors. The simulation includes generating random signals and fading gains, and implementing the transmitter with the encoder and the receiver with the decoder for both fixed-rate and rateless coded schemes. The optimal performance of the systems will be determined by varying different system parameters. Concurrently, the student will be instructed in the relevant statistical communication theory and coding theory, and will formulate mathematical models for the fixed-rate and rateless coded protocols and for coding and decoding at the transmitters and receivers, respectively.

The ultimate goal is to complete both mathematical and simulation comparisons between the fixed-rate and rateless coded relaying networks. If progress is rapid, the important effect of optimal power allocation between different users will be investigated via analytical optimization methods and simulation.

Lakshmi Sumala
Faculty Supervisor: 
Dr. Norman Beaulieu