Deep Ensemble of Weighted Viterbi Decoders for Tail-Biting Convolutional Codes

Tail-biting convolutional codes extend the classical zero-termination convolutional codes: Both encoding schemes force the equality of start and end states, but under the tail-biting each state is a valid termination. This paper proposes a machine learning approach to improve the state-of-the-art decoding of tail-biting codes, focusing on the widely employed short length regime as in the LTE standard. This standard also includes a CRC code. First, we parameterize the circular Viterbi algorithm, a baseline decoder that exploits the circular nature of the underlying trellis. An ensemble combines multiple such weighted decoders, and each decoder specializes in decoding words from a specific region of the channel words’ distribution. A region corresponds to a subset of termination states; the ensemble covers the entire states space. A non-learnable gating satisfies two goals: it filters easily decoded words and mitigates the overhead of executing multiple weighted decoders. The CRC criterion is employed to choose only a subset of experts for decoding purpose. Our method achieves FER improvement of up to 0.75 dB over the CVA in the waterfall region for multiple code lengths, adding negligible computational complexity compared to the circular Viterbi algorithm in high signal-to-noise ratios (SNRs).

High-Speed Viterbi Decoder

The Viterbi algorithm is the most well-known trellis-based maximum likelihood decoding algorithm. Trellis decoding is used to recover encoded information that was corrupted during transmission over a noisy channel. The Viterbi algorithm is implemented with a Viterbi decoder. High-speed applications require high-speed Viterbi decoders. Therefore, many hardware solutions have been proposed to improve the performance of Viterbi decoders. These hardware solutions explore the properties of the Viterbi algorithm to simplify and improve the architecture of the decoder. In particular, statistical properties of the algorithm are used to design parallel Viterbi decoders with very high data decoding rates. The article focuses on the implementation of high-speed Viterbi decoders.

Deep Ensemble of Weighted Viterbi Decoders for Tail-Biting Convolutional Codes

Tail-biting convolutional codes extend the classical zero-termination convolutional codes: Both encoding schemes force the equality of start and end states, but under the tail-biting each state is a valid termination. This paper proposes a machine-learning approach to improve the state-of-the-art decoding of tail-biting codes, focusing on the widely employed short length regime as in the LTE standard. This standard also includes a CRC code. First, we parameterize the circular Viterbi algorithm, a baseline decoder that exploits the circular nature of the underlying trellis. An ensemble combines multiple such weighted decoders, each decoder specializes in decoding words from a specific region of the channel words' distribution. A region corresponds to a subset of termination states; the ensemble covers the entire states space. A non-learnable gating satisfies two goals: it filters easily decoded words and mitigates the overhead of executing multiple weighted decoders. The CRC criterion is employed to choose only a subset of experts for decoding purpose. Our method achieves FER improvement of up to 0.75dB over the CVA in the waterfall region for multiple code lengths, adding negligible computational complexity compared to the circular Viterbi algorithm in high SNRs.

Viterbi Decoder in Hardware

Trellis decoding is used to recover encoded information that was corrupted during transmission over a noisy channel. The Viterbi algorithm is the most well-known trellis-based maximum likelihood decoding algorithm. The Viterbi algorithm is executed by a Viterbi decoder. Different hardware solutions may be considered to implement a Viterbi decoder with different design requirements in terms of area, performance, power consumption, among others. The most appropriate solution depends on the metric requirements of the application as well as on the target technology. Properties of the Viterbi algorithm are used to simplify and improve the architecture of the Viterbi decoder. In particular, statistical properties of the Viterbi algorithm are used to design parallel Viterbi decoders with very high data decoding rates. The chapter focuses on the implementation of a Viterbi decoder in hardware, including optimizations to improve the area and performance.

Viterbi Decoder in Hardware

Trellis decoding is used to recover encoded information that was corrupted during transmission over a noisy channel. The Viterbi algorithm is the most well known trellis-based maximum likelihood decoding algorithm. The Viterbi algorithm is executed by a Viterbi decoder. Different hardware solutions may be considered to implement a Viterbi decoder with different design requirements in terms of area, performance, power consumption, among others. The most appropriate solution depends on the metric requirements of the application as well as on the target technology. Properties of the Viterbi algorithm are used to simplify and improve the architecture of the Viterbi decoder. In particular, statistical properties of the Viterbi algorithm are used to design parallel Viterbi decoders with very high data decoding rates. The article focuses on the implementation of a Viterbi decoder in hardware, including optimizations to improve the area and performance.