IEEE Communications Letters · 2018 · 23 citations · 11 references
EngineeringComputer EngineeringComputer ScienceAlgorithmsMachine LearningOptimizationChannel CodingCoding TheoryDigital Signal ProcessingIterative SolversParameter EstimationComputational Complexity Of AlgorithmsDecoding AlgorithmsIterative DecodingIterative Design ProcessIterative SystemsTurbo Codes
A new parameter estimation algorithm is proposed for a recursive systematic convolutional, or sub-codes of turbo codes. The method is based on optimized cosinoidal cost function of RSC encoder coefficients. A simple iterative process by gradient ascent is designed for the maximum of the cosinoidal cost function. The simulation results show that the proposed method can improve the performance significantly with lower computational complexity, compared to existing algorithms.
11