Tuned turbo codes

Abstract

As implied by previous studies, there exists a fundamental trade-off between the minimum distance and the iterative decoding convergence behavior of a turbo code. While capacity achieving code ensembles typically are asymptotically bad in the sense that their minimum distance does not grow linearly with block length and they therefore exhibit an error floor at medium to high signal to noise ratios, asymptotically good codes usually converge further away from channel capacity. In this paper we present so-called tuned turbo codes, a family of asymptotically good hybrid concatenated code ensembles, where minimum distance growths and convergence thresholds can be traded-off using a tuning parameter lambda. By decreasing lambda, the asymptotic minimum distance growth rate coefficient is reduced for the sake of improved iterative decoding convergence behavior, and thus the code performance can be tuned to fit the desired application.