The capacity of average power constrained additive non-Gaussian noise channels
Jihad Fahs, Nizar Ajeeb, Ibrahim Abou‐Faycal
Abstract
Jihad Fahs, Nizar Ajeeb, Ibrahim Abou‐Faycal
Abstract
It is well known that a Gaussian input achieves the capacity of the linear additive white Gaussian noise channel under the average power constraint. While the continuity of an optimal input of a continuous channel might be expected, the Gaussian noise presents the only such scenario. In fact, we study in this paper the capacity-achieving inputs for the linear additive noise channel where the noise is not necessarily Gaussian. We impose an average power constraint on the input and prove that, except for a Gaussian channel, the optimal input is of a discrete nature.
OpenAlex reports 11 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
It is well known that a Gaussian input achieves the capacity of the linear additive white Gaussian noise channel under the average power constraint. While the continuity of an optimal input of a continuous channel might be expected, the Gaussian noise presents the only such scenario. In fact, we study in this paper the capacity-achieving inputs for the linear additive noise channel where the noise is not necessarily Gaussian. We impose an average power constraint on the input and prove that, except for a Gaussian channel, the optimal input is of a discrete nature.
Key concepts: Additive white Gaussian noise, Gaussian noise, Shannon–Hartley theorem, Gaussian, Noise (video), Channel (broadcasting), Constraint (computer-aided design), Gaussian random field