Abstract
Massive multiple-input multiple-output (MIMO) attracts considerable interest by increasing the spectral efficiency, and which may be adopted in future communication system. Both base station (BS) and user equipment (UE) can be equipped with large-scale antennas using frequency division duplex (FDD) schemes. The major obstacle that reduces the performance is the overhead due to the reference signal. Moreover, the channel state information (CSI) of the uplink (UL) channel cannot be simply used for downlink (DL) pre-coding. In this paper, we focus on reconstructing wireless channels on a frequency band by observing the channel response on neighboring frequency band. The inference of DL channel is conducted by utilizing parameters of multipath components extracted from UL observations, using high resolution estimation algorithms, e.g. the Space-Alternating Generalized Expectation-maximization (SAGE). Four calibration methods are proposed to improve the quality of the inferred DL channel response. Those four methods take into account the channel composition based on the availability of the initial channel observations. Their performance is demonstrated by using real massive MIMO measurements. The results show these calibration methods have their individual application scenarios and all outperform in Line-of-Sight (LoS) scenarios. These calibration methods prove a methodological way to improve the correlation of neighboring channel.
Highlights
Massive Multiple-input multiple-output (MIMO) technology has been widely used in wireless communications because it improves the spectral efficiency and link reliability by using diversity techniques
The performance of frequency division duplex (FDD) systems can be significantly reduced due to the errors in quantization of the feedback information compared to the time division duplex (TDD) system [7]
Comparing the performance and complexity of the existing method, a novel FDD massive MIMO system based on a spatial DL channel estimation scheme is proposed in [26], [27]and the results showed significant performance improvement
Summary
Massive Multiple-input multiple-output (MIMO) technology has been widely used in wireless communications because it improves the spectral efficiency and link reliability by using diversity techniques. Based on the obtained results, four calibration methods are proposed to improve the quality of the inferred DL channel response. The UL channel observations are processed and the High-Resolution Parameter Estimation (HRPE) method utilizing the SAGE algorithm [22], [23] is used to estimate the parameters of channel multipath components (MPCs) based on a planar wave assumption With these estimation results, the DL channel frequency responses are inferred and compared with the DL channel frequency responses observed. Comparing the performance and complexity of the existing method, a novel FDD massive MIMO system based on a spatial DL channel estimation scheme is proposed in [26], [27]and the results showed significant performance improvement.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have