Viterbi Detection for Compressively Sampled FHSS-GFSK Signals

Lu Pan, Michael W Marcellin, William E. Ryan, Bane V Vasic

Research output: Contribution to journalArticle

2 Citations (Scopus)

Abstract

This paper proposes a sequence detector designed to retrieve data bits from compressively sampled frequency-hopping spread spectrum (FHSS) Gaussian frequency-shift keying (GFSK) signals. The received signal waveform is not reconstructed from the compressive measurements, nor are received bits detected on a symbol-by-symbol basis. Rather, the entire sequence of transmitted symbols is detected from the entire sequence of compressive measurements. Another novel aspect of the work is that a non-cooperative scenario is assumed. Specifically, the receiver is assumed to have no prior knowledge of the specific spread spectrum hopping sequence used by the transmitter. The most significant contribution of the work is the design of adaptive sampling kernels that exploit the structure of an FHSS-GFSK signal to obtain significant performance improvements over random-kernel sampling. The resulting system can automatically choose an appropriate compression ratio as a function of the signal-to-noise ratio (SNR) without explicit knowledge of the SNR. Additionally, the noise folding problem present in random-kernel sampling is greatly alleviated by use of the adaptive sampling kernels. Compared with Nyquist sampling, adaptive compressive sampling offers compression ratios ranging from 20 to 32, depending on the SNR while suffering less than 1 dB loss in the resulting bit error rate.

Original languageEnglish (US)
Article number7169602
Pages (from-to)5965-5975
Number of pages11
JournalIEEE Transactions on Signal Processing
Volume63
Issue number22
DOIs
StatePublished - Nov 15 2015

Fingerprint

Frequency hopping
Frequency shift keying
Sampling
Signal to noise ratio
Bit error rate
Transmitters
Detectors

Keywords

  • Adaptive sampling
  • compressive sampling
  • Viterbi detection

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing

Cite this

Viterbi Detection for Compressively Sampled FHSS-GFSK Signals. / Pan, Lu; Marcellin, Michael W; Ryan, William E.; Vasic, Bane V.

In: IEEE Transactions on Signal Processing, Vol. 63, No. 22, 7169602, 15.11.2015, p. 5965-5975.

Research output: Contribution to journalArticle

@article{680c445da96c4c45a8ebfce662a3ddac,
title = "Viterbi Detection for Compressively Sampled FHSS-GFSK Signals",
abstract = "This paper proposes a sequence detector designed to retrieve data bits from compressively sampled frequency-hopping spread spectrum (FHSS) Gaussian frequency-shift keying (GFSK) signals. The received signal waveform is not reconstructed from the compressive measurements, nor are received bits detected on a symbol-by-symbol basis. Rather, the entire sequence of transmitted symbols is detected from the entire sequence of compressive measurements. Another novel aspect of the work is that a non-cooperative scenario is assumed. Specifically, the receiver is assumed to have no prior knowledge of the specific spread spectrum hopping sequence used by the transmitter. The most significant contribution of the work is the design of adaptive sampling kernels that exploit the structure of an FHSS-GFSK signal to obtain significant performance improvements over random-kernel sampling. The resulting system can automatically choose an appropriate compression ratio as a function of the signal-to-noise ratio (SNR) without explicit knowledge of the SNR. Additionally, the noise folding problem present in random-kernel sampling is greatly alleviated by use of the adaptive sampling kernels. Compared with Nyquist sampling, adaptive compressive sampling offers compression ratios ranging from 20 to 32, depending on the SNR while suffering less than 1 dB loss in the resulting bit error rate.",
keywords = "Adaptive sampling, compressive sampling, Viterbi detection",
author = "Lu Pan and Marcellin, {Michael W} and Ryan, {William E.} and Vasic, {Bane V}",
year = "2015",
month = "11",
day = "15",
doi = "10.1109/TSP.2015.2461518",
language = "English (US)",
volume = "63",
pages = "5965--5975",
journal = "IEEE Transactions on Signal Processing",
issn = "1053-587X",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "22",

}

TY - JOUR

T1 - Viterbi Detection for Compressively Sampled FHSS-GFSK Signals

AU - Pan, Lu

AU - Marcellin, Michael W

AU - Ryan, William E.

AU - Vasic, Bane V

PY - 2015/11/15

Y1 - 2015/11/15

N2 - This paper proposes a sequence detector designed to retrieve data bits from compressively sampled frequency-hopping spread spectrum (FHSS) Gaussian frequency-shift keying (GFSK) signals. The received signal waveform is not reconstructed from the compressive measurements, nor are received bits detected on a symbol-by-symbol basis. Rather, the entire sequence of transmitted symbols is detected from the entire sequence of compressive measurements. Another novel aspect of the work is that a non-cooperative scenario is assumed. Specifically, the receiver is assumed to have no prior knowledge of the specific spread spectrum hopping sequence used by the transmitter. The most significant contribution of the work is the design of adaptive sampling kernels that exploit the structure of an FHSS-GFSK signal to obtain significant performance improvements over random-kernel sampling. The resulting system can automatically choose an appropriate compression ratio as a function of the signal-to-noise ratio (SNR) without explicit knowledge of the SNR. Additionally, the noise folding problem present in random-kernel sampling is greatly alleviated by use of the adaptive sampling kernels. Compared with Nyquist sampling, adaptive compressive sampling offers compression ratios ranging from 20 to 32, depending on the SNR while suffering less than 1 dB loss in the resulting bit error rate.

AB - This paper proposes a sequence detector designed to retrieve data bits from compressively sampled frequency-hopping spread spectrum (FHSS) Gaussian frequency-shift keying (GFSK) signals. The received signal waveform is not reconstructed from the compressive measurements, nor are received bits detected on a symbol-by-symbol basis. Rather, the entire sequence of transmitted symbols is detected from the entire sequence of compressive measurements. Another novel aspect of the work is that a non-cooperative scenario is assumed. Specifically, the receiver is assumed to have no prior knowledge of the specific spread spectrum hopping sequence used by the transmitter. The most significant contribution of the work is the design of adaptive sampling kernels that exploit the structure of an FHSS-GFSK signal to obtain significant performance improvements over random-kernel sampling. The resulting system can automatically choose an appropriate compression ratio as a function of the signal-to-noise ratio (SNR) without explicit knowledge of the SNR. Additionally, the noise folding problem present in random-kernel sampling is greatly alleviated by use of the adaptive sampling kernels. Compared with Nyquist sampling, adaptive compressive sampling offers compression ratios ranging from 20 to 32, depending on the SNR while suffering less than 1 dB loss in the resulting bit error rate.

KW - Adaptive sampling

KW - compressive sampling

KW - Viterbi detection

UR - http://www.scopus.com/inward/record.url?scp=84959421587&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84959421587&partnerID=8YFLogxK

U2 - 10.1109/TSP.2015.2461518

DO - 10.1109/TSP.2015.2461518

M3 - Article

AN - SCOPUS:84959421587

VL - 63

SP - 5965

EP - 5975

JO - IEEE Transactions on Signal Processing

JF - IEEE Transactions on Signal Processing

SN - 1053-587X

IS - 22

M1 - 7169602

ER -