Statistical CSI Driven Transmit Antenna Selection and Power Adaptation in Underlay Spectrum Sharing Systems

• Published in: IEEE transactions on communications Vol. 69; no. 5; pp. 2923 - 2934
• Main Authors: Sarvendranath, Rimalapudi, Mehta, Neelesh B.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01-05-2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptation; Adaptation models; Adaptive arrays; antenna selection; Antennas; channel state information; Interference; Interference constraints; power adaptation; Radio frequency; Spectrum sharing; Statistical analysis; Stochastic processes; Transmitting antennas; underlay; Wireless communication
• ISSN: 0090-6778; 1558-0857
• DOI: 10.1109/TCOMM.2021.3057871

In underlay spectrum sharing, transmit antenna selection (TAS) improves the performance of a secondary system and helps it control the interference it causes to a primary system. TAS does so with a hardware complexity and cost comparable to a single antenna system. We present a novel and optimal joint TAS and continuous power adaptation rule for a practically relevant, less explored model in which the secondary transmitter knows only the statistics of channel gains from itself to one or more primary receivers. The rule minimizes the average symbol error probability (SEP) of the secondary system for an entire class of stochastic interference constraints. This general class subsumes the average interference constraint and its novel generalization, and the interference-outage constraint. We derive closed-form expressions for the transmit power and selected antenna. We then develop a general analysis of the optimal average SEP that applies to several widely-used fading models. We also present computationally-efficient approaches to determine the parameters that specify the optimal rule. Our comprehensive numerical results characterize the very different impacts of the interference constraint on both secondary and primary systems. They show that the optimal rule reduces the average SEP by two orders of magnitude compared to conventional approaches.