The aim of this thesis is to study the impact of different decoding algorithms on the performance of error correcting codes proposed to update the telecommand link standard for space missions. Specifically, we refer to binary and non-binary Low Density Parity Check Codes (LDPC). To evaluate the performance we consider three different channel models of great interest in space communications, namely: Additive White Gaussian Noise (AWGN), AWGN with pulsed jamming and AWGN with solar scintillation; and three different metrics, namely, Codeword Error Rate (CER), Undetected Codeword Error Rate (UCER) and complexity. The list of the considered decoding algorithms includes iterative decoders, like: the Sum Product Algorithm (SPA), the Min-Sum (MS) and the Normal- ized Min-Sum (NMS) algorithms; but also a non-iterative procedure, namely the Most Reliable Basis (MRB) algorithm. For the non-binary codes, in turn, we consider a specific implementation of the Belief Propagation (BP) method. Besides these classical decoding schemes, we also discuss an alternative approach, called hybrid decoder. The hybrid decoder is constituted by the parallel combination of an iterative decoder and a reliability based decoder. We show that this scheme is able to provide some advantage with respect to classical decoding algorithms. In particular, the hybrid decoder combines the excellent error rate performance of the MRB decoder with an average complexity only slightly higher than that of the iterative algorithms. Most of the work is developed through theoretical analyses and numerical simulations. However, we also provide some preliminary considerations on the practical implementation of the most efficient decoding algorithms. In particular, we discuss the not yet addressed problem of implementing the MRB algorithm on Field Programmable Gate Array (FPGA) by proposing, as possible solution, a combined software-hardware implementation.
Lo scopo di questa tesi ´e studiare l’impatto di diversi algoritmi di decodifica sulle performance dei codici a correzione d’errore proposti per l’aggiornamento dello standard del telocommand link nelle missioni spaziali, che sono rappresentati dai codici Low Density Parity Check (LDPC) binari e non binari. Per la valutazione delle performance si sono considerati tre diversi modelli di canale di grande rilevanza nello comunicazioni con lo spazio, ossia: Additive White Gaussian Noise (AWGN), canale AWGN con pulsed jamming e canale AWGN con scintillazione solare; e tre diverse metriche, ossia: Codeword Error Rate (CER), Undetected Codeword Error Rate (UCER) e complessità. La lista degli algoritmi considerati include algoritmi iterativi, come: il Sum Product Algorithm (SPA), l’algoritmo Min-Sum (MS) e l’algoritmo Normalized Min-Sum (NMS); ma anche una procedura non iterativa, ossia l’algortimo Most Reliable Basis (MRB). Infine per i codici non binari consideriamo una specifica implementazione del metodo Belief Propagation (BP). Oltre a questi schemi di decodifica classici, viene proposto anche un approccio alternativo, chiamato decoder ibrido. Il decoder ibrido è costituito dalla combinazione parallela di un algoritmo iterativo e di un algoritmo basato sull’affidabilità. Vengono mostrati i vantaggi che questo schema ´è in grado di fornire rispetto agli algoritmi di decodifica classici. Nello specifico, il decoder ibrido combina le eccellenti performance in termini di error rate dell’algoritmo MRB con una complessità media lievemente maggiore di quella degli algoritmi iterativi. Gran parte del lavoro è stato sviluppato tramite analisi teorica e simulazioni. Tuttavia, vengono anche fornite alcune considerazioni preliminari sull’implementazione pratica dei più efficienti algoritmi di decodifica. In par- ticolare, viene discussa la realizzazione dell’algoritmo MRB su Field Programmable Gate Array (FPGA) proponendo, come possibile soluzione, una implementazione combinata hardware-software.
Short codes for telecommand space link / Maturo, Nicola. - (2016 Feb 26).
Short codes for telecommand space link
Maturo, Nicola
2016-02-26
Abstract
The aim of this thesis is to study the impact of different decoding algorithms on the performance of error correcting codes proposed to update the telecommand link standard for space missions. Specifically, we refer to binary and non-binary Low Density Parity Check Codes (LDPC). To evaluate the performance we consider three different channel models of great interest in space communications, namely: Additive White Gaussian Noise (AWGN), AWGN with pulsed jamming and AWGN with solar scintillation; and three different metrics, namely, Codeword Error Rate (CER), Undetected Codeword Error Rate (UCER) and complexity. The list of the considered decoding algorithms includes iterative decoders, like: the Sum Product Algorithm (SPA), the Min-Sum (MS) and the Normal- ized Min-Sum (NMS) algorithms; but also a non-iterative procedure, namely the Most Reliable Basis (MRB) algorithm. For the non-binary codes, in turn, we consider a specific implementation of the Belief Propagation (BP) method. Besides these classical decoding schemes, we also discuss an alternative approach, called hybrid decoder. The hybrid decoder is constituted by the parallel combination of an iterative decoder and a reliability based decoder. We show that this scheme is able to provide some advantage with respect to classical decoding algorithms. In particular, the hybrid decoder combines the excellent error rate performance of the MRB decoder with an average complexity only slightly higher than that of the iterative algorithms. Most of the work is developed through theoretical analyses and numerical simulations. However, we also provide some preliminary considerations on the practical implementation of the most efficient decoding algorithms. In particular, we discuss the not yet addressed problem of implementing the MRB algorithm on Field Programmable Gate Array (FPGA) by proposing, as possible solution, a combined software-hardware implementation.File | Dimensione | Formato | |
---|---|---|---|
Tesi_Maturo.pdf
Solo gestori archivio
Tipologia:
Tesi di dottorato
Licenza d'uso:
Non specificato
Dimensione
3.42 MB
Formato
Adobe PDF
|
3.42 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.