Abstract
This paper proposed a configurable sub-connected architecture with a framework that dynamically activates the near-optimal subset of antennas and RF chains to implement energy-efficient hybrid precoding in millimeter wave multiple-input multiple-output system. Since the exhaust search is computational intractable, we propose a two-stage hybrid precoding algorithm, where the digital precoder is designed to eliminate the inter-user interference by zero-forcing rule. Specifically, in the first stage, we introduce an extended cross-entropy algorithm that adaptively updates the probability distribution of potential states in the analog precoder matrix, which can generate a solution that is close to the optimal with a sufficiently high probability. In the second stage, a QR-based subset selection algorithm is proposed to pick the near-optimal subset of the RF chains to further cut down on the energy cost. Simulation results show that proposed extend-CE algorithm gets favorable performance in terms of energy efficiency, and QR-based RF chain selection can achieve a near-optimal performance. Other hybrid precoding algorithms can also be incorporated into the proposed RF chain selection algorithm.
Highlights
Millimeter-wave massive multiple-input multiple-output (MIMO) has been regarded as a key technology to meet the increasing traffic and critical energy efficiency demand in cellular network [1][2]
SIMULATION RESULTS the performance of our proposed precoding algorithm is evaluated in the downlink mmWave MIMO systems
In this paper, a sub-connected hybrid precoding structure with configurable digital and analog connections has been proposed for the mmWave MU-MIMO downlink transmission
Summary
Millimeter-wave (mmWave) massive multiple-input multiple-output (MIMO) has been regarded as a key technology to meet the increasing traffic and critical energy efficiency demand in cellular network [1][2]. The power consumption model investigated in [32] has been widely used for analyzing the energy efficiency optimization problem Based on this model, many energy-efficient hybrid precoding schemes have been proposed in mmWave MIMO cellular communications [33,34,35,36,37,38,39,40,41,42,43]. The baseband channel matrix should have a sufficient rank to transmit enough streams to support all the clients, while at the same time, the number of active RF chains should be as less as possible to save energy consumption This problem can be derived as identifying the most representative columns of the equivalent analog channel matrix. IN is the N×N identity matrix, and 0 is the all-zero matrix
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have