1995
DOI: 10.1109/26.481226
|View full text |Cite
|
Sign up to set email alerts
|

Maximum likelihood decoding of uncoded and coded PSK signal sequences transmitted over Rayleigh flat-fading channels

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
51
0

Year Published

1998
1998
2013
2013

Publication Types

Select...
6
2

Relationship

0
8

Authors

Journals

citations
Cited by 98 publications
(51 citation statements)
references
References 25 publications
0
51
0
Order By: Relevance
“…Hence, for increasing values of C, the number of factor nodes increases linearly but the computational burden at each factor node remains the same. Hence, the algorithm complexity is linear in C. This is a fundamental difference with respect to trellis-based linear predictive receivers [42], [43], [71], [73], [75], whose complexity is exponential in the prediction order, and suggests that by using the tool represented by FGs and SPA new computationally efficient algorithms can be derived. In addition, in this modified FG there are no cycles of length 4 in the part of the graph modeling the channel and, by adopting the flooding schedule, the derived algorithm is also well suited for a fully parallel implementation of the detector/ decoder.…”
Section: ) Examplevequal Energy Signals Transmitted Over a Flatmentioning
confidence: 99%
See 2 more Smart Citations
“…Hence, for increasing values of C, the number of factor nodes increases linearly but the computational burden at each factor node remains the same. Hence, the algorithm complexity is linear in C. This is a fundamental difference with respect to trellis-based linear predictive receivers [42], [43], [71], [73], [75], whose complexity is exponential in the prediction order, and suggests that by using the tool represented by FGs and SPA new computationally efficient algorithms can be derived. In addition, in this modified FG there are no cycles of length 4 in the part of the graph modeling the channel and, by adopting the flooding schedule, the derived algorithm is also well suited for a fully parallel implementation of the detector/ decoder.…”
Section: ) Examplevequal Energy Signals Transmitted Over a Flatmentioning
confidence: 99%
“…where the order of Markovianity C can be interpreted as the prediction order, fp i g C i¼1 are the prediction coefficients (which depend on state S k , but not on symbol a k ), and 2 k represents the mean square prediction error at epoch k. The result in (56), which can be derived from (53) owing to the Gaussianity of the observable, was obtained in [42], [43], and [71]- [75] as a solution for maximum likelihood sequence detection over fading channels. Ì…”
Section: Finite-memory Iterative Detectionmentioning
confidence: 99%
See 1 more Smart Citation
“…We observed that this can lead to numerical problems at high SNR, where the noise variance becomes small. To overcome this problem, one may increase the accuracy of the numerical integration techniques used to compute (13), or prevent the variances of the Gaussian pdfs to become too small and trigger numerical problems.…”
Section: B Multi-trellis Siso Algorithm For Fading Channelmentioning
confidence: 99%
“…The basic idea is to extend the observation length to be larger than two consecutive signaling intervals and to gain additional information among received signals to improve detection. In time-varying channels, these receivers are closely related to a linear prediction receiver (e.g., [9], [12], and [13]). The linear prediction receiver finds the best transmitted sequence that minimizes Euclidean distance between the channel gains formed by the received signals and those from the prediction.…”
Section: Introductionmentioning
confidence: 99%