2013
DOI: 10.1364/jocn.5.000329
|View full text |Cite
|
Sign up to set email alerts
|

Capacity and Nonuniform Signaling for Discrete-Time Poisson Channels

Abstract: The Poisson photon-counting model is accurate for optical channels with low received intensity, such as long-range intersatellite optical wireless links. This work considers the computation of the channel capacity and the design of capacity-approaching, nonuniform signaling for discrete-time Poisson channels in the presence of dark current and underaverage and peak amplitude constraints. Although the capacity of this channel is unknown, numerical computation of the channel capacity is implemented using a parti… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
8
0

Year Published

2014
2014
2022
2022

Publication Types

Select...
5
2

Relationship

2
5

Authors

Journals

citations
Cited by 13 publications
(8 citation statements)
references
References 14 publications
0
8
0
Order By: Relevance
“…In this manner, the shot-noise distorted Rx is transformed to a Rx which is affected by a white Gaussian signal-independent noise. Using this simple transformation, we can avoid non-uniform signaling that needs rather complex optimization for calculating the signal constellation [33], [34]. As shown in the block diagram of Fig.…”
Section: B Optimal Pam Signalingmentioning
confidence: 99%
“…In this manner, the shot-noise distorted Rx is transformed to a Rx which is affected by a white Gaussian signal-independent noise. Using this simple transformation, we can avoid non-uniform signaling that needs rather complex optimization for calculating the signal constellation [33], [34]. As shown in the block diagram of Fig.…”
Section: B Optimal Pam Signalingmentioning
confidence: 99%
“…4) It then finds the most likely choice of c nK 2 2 , denoted byĉ nK 2 2 , based onŵ n 1 ,ŵ n 2 ,û n 3 , and c nK 2 2 H ′ 2 (which can be deduced from c nK 2 2Ĥ ′ 2 and the fact that c nK 2 2H ′ 2 is a zero vector). This can be realized via joint demapping and decoding with (H ′ 2 , φ 2 ) defining the factor graph and (ŵ n 1 ,ŵ n 2 ,û n 3 ) serving as the channel output (see, e.g., [38] (which can be deduced from c nK 1 1Ĥ1 and the fact that c nK 1 1H1 is a zero vector). This can be realized via joint demapping and decoding with (H 1 , φ 1 ) defining the factor graph and (ŵ n 1 ,û n 2 ,û n 3 ) serving as the channel output.…”
Section: Codebook Generationmentioning
confidence: 99%
“…In this section we generalize the approximation scheme introduced in Section 2 to memoryless channels with continuous input and countable output alphabets. The class of discrete-time Poisson channels is an example of such channels with particular interest in applications, for example to model direct detection optical communication systems [10,26,27]. Consider X ⊆ R as the input alphabet set and Y = N 0 as the output alphabet set.…”
Section: Channels With Continuous Input and Countable Output Alphabetsmentioning
confidence: 99%
“…Based on sequential Monte-Carlo integration methods (a.k.a. particle filters), the Blahut-Arimoto algorithm has been extended to memoryless channels with continuous input and output alphabets [9][10][11][12]. As shown in several examples, this approach seems to be powerful in practice, however a rate of convergence has not been proven.…”
Section: Introductionmentioning
confidence: 99%