On the Capacity-Achieving Scheme and Capacity of 1-Bit ADC Gaussian-Mixture Channels

Rahman, Md Hasan and Ranjbar, Mohammad and Tran, Nghi H. (2020) On the Capacity-Achieving Scheme and Capacity of 1-Bit ADC Gaussian-Mixture Channels. EAI Endorsed Transactions on Industrial Networks and Intelligent Systems, 7 (22): 4. ISSN 2410-0218

[thumbnail of PDF]
Text (PDF)
eai.31-1-2020.162830.pdf - Published Version
Available under License Creative Commons Attribution No Derivatives.

Download (2MB) | Preview


This paper addresses the optimal signaling scheme and capacity of an additive Gaussian mixture (GM) noise channel using 1-bit analog-to-digital converters (ADCs). The consideration of GM noise provides a more realistic baseline for the analysis and design of co-channel interference links and networks. Towards that goal, we first show that the capacityachieving input signal is π/2 circularly symmetric. By examining a necessary and sufficient Kuhn–Tucker condition (KTC) for an input to be optimal, we demonstrate that the maximum number of optimal mass points is four. Our proof relies on Dubin’s theorem and the fact that the KTC coefficient is positive, i.e., the power constraint is active. By combining with the π/2 circularly symmetric property, it is then concluded the optimal input is unique, and it has exactly four mass points forming a square centered at the origin. By further checking the first and second derivatives of the modified KTC, it is then shown that the phase of the optimal mass point located in the first quadrant is π/4. Thus, the capacity-achieving input signal is QPSK. This result helps us obtain the channel capacity in closed-form.

Item Type: Article
Uncontrolled Keywords: 1-bit ADC, Capacity, Gaussian-Mixture, Kuhn-Tucker Condition, Mutual Information
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
QA75 Electronic computers. Computer science
Depositing User: EAI Editor I.
Date Deposited: 11 Sep 2020 07:53
Last Modified: 14 Sep 2020 15:17
URI: https://eprints.eudl.eu/id/eprint/228

Actions (login required)

View Item
View Item