A Dynamical System View for the Turbo Decoding Algorithm Based on the Gaussian Approximation
نویسندگان
چکیده
The Gaussian approximation has been used to study iterative decoding algorithms of graph-based codes like turbo codes [7, 8] and low-density parity check (LDPC) codes [6]. In this work, the Gaussian approximation is used to describe the turbo decoding algorithm (TDA) as an elegant discrete dynamical system and explicit iterative equations for the evolution of the BER with the decoder phases of the TDA are presented. This formulation is used to study the SNR thresholds for turbo codes and the fixed points of the TDA. The issues of stability of TDA’s fixed points for SNRs below the threshold and rate of convergence of the TDA to very low BERs for SNRs above the threshold are also addressed.
منابع مشابه
Iterative turbo decoder analysis based on density evolution
We track the density of extrinsic information in iterative turbo decoders by actual density evolution, and also approximate it by consistent Gaussian density functions. The approximate model is verified by experimental measurements. We view the evolution of these density functions through an iterative decoder as a nonlinear dynamical system with feedback. Iterative decoding of turbo codes and o...
متن کاملBifurcations and Chaos in Turbo Decoding Algorithms
The turbo decoding algorithm is a high-dimensional dynamical system parameterized by a large number of parameters (for a practical realization the turbo decoding algorithm has more than 10 variables and is parameterized by more than 10 parameters). In this chapter we treat the turbo decoding algorithm as a dynamical system parameterized by a single parameter that closely approximates the signal...
متن کاملA Flexible Link Radar Control Based on Type-2 Fuzzy Systems
An adaptive neuro fuzzy inference system based on interval Gaussian type-2 fuzzy sets in the antecedent part and Gaussian type-1 fuzzy sets as coefficients of linear combination of input variables in the consequent part is presented in this paper. The capability of the proposed method (we named ANFIS2) for function approximation and dynamical system identification is remarkable. The structure o...
متن کاملA novel stopping criterion for turbo codes based on the average a posteriori entropy
In this paper, we treat the turbo decoding algorithm as a dynamical system parameterized by a single parameter that closely approximates the signal-to-noise ratio (SNR). A whole range of phenomena known to occur in nonlinear systems, like the existence of multiple fixed points, oscillatory behavior, bifurcations, chaos and transient chaos are found in the turbo decoding algorithm. As an applica...
متن کاملAn Iterative Decoding Algorithm for Channels with Additive Linear Dynamical Noise
In this paper, an iterative decoding algorithm for channels with additive linear dynamical noise is presented. The proposed algorithm is based on the tightly coupled two inference algorithms: the sum-product algorithm which infers the information symbols of an low density parity check(LDPC) code and the Kalman smoothing algorithm which infers the channel states. The linear dynamical noise are t...
متن کامل