Jianzhi Li

On Indoor Millimeter Wave Massive MIMO Channels: Measurement and Simulation

The millimeter wave (mmWave) communications and massive multiple-input multiple-output (MIMO) are both widely considered to be the candidate technologies for the fifth generation mobile communication system. It is thus a good idea to combine these two technologies to achieve a better performance for large capacity and high data-rate transmission. However, one of the fundamental challenges is the characterization of mmWave massive MIMO channel. Most of the previous investigations in mmWave channel only focus on single-input single-output links or MIMO links, whereas the research of massive MIMO channels mainly focus on a frequency band below 6 GHz. This paper investigates the channel behaviors of massive MIMO at a mmWave frequency band around 26 GHz. An indoor mmWave massive MIMO channel measurement campaign with 64 and 128 array elements is conducted, based on which, path loss, shadow fading, root-mean-square (RMS) delay spread, and coherence bandwidth are extracted. Then, by using our developed ray-tracing simulator calibrated by the measurement data, we make the extensive ray-tracing simulations with 1024 antenna elements in the same indoor scenario, and get insights into the variation tendency of mean delay and the RMS delay with different array elements. It is observed that the measurement and the ray-tracing-based simulation results have reached a good agreement.

Measurement-Based Characterizations of Indoor Massive MIMO Channels at 2 GHz, 4 GHz, and 6 GHz Frequency Bands

Massive MIMO has been chosen as one of the candidate technologies of the fifth-generation mobile communication system (5G), and channel modeling of massive MIMO is of great importance. The most direct and effective approach to investigate the propagation characteristics of massive MIMO channels is channel measurements. However, there are only few measurements of massive MIMO channels, and there still lacks deep investigations of massive MIMO channel characteristics. In this paper, we present a measurement campaign of indoor massive MIMO channels, by using a linear large-scale array with 64 elements. The measurements are conducted at 2 GHz, 4 GHz, and 6 GHz, respectively, with a bandwidth of 200 MHz. Both LOS and NLOS propagation scenarios are considered in the measurements. The basic channel parameters are extracted, including path loss, delay spread, and coherence bandwidth. The non-stationarity of radio channels, which is reflected by the variations of delay spread and coherence bandwidth over different array locations, is discussed. The impact of carrier frequency on the above channel parameters is further discussed. The results would be useful for the design of massive MIMO system in the indoor environments.

Parameter estimation using SAGE algorithm based on Massive MIMO channel measurements

This paper presents the channel measurements for Massive MIMO communications in indoor scenario at 6 GHz. The channel parameters are jointly estimated using sliding-window SAGE algorithm over the antenna array. The estimated results in delay and angular domains are presented. Considering the large size of antenna array at the transmitter side, the non-stationary phenomenon in the space domain is observed.

Directional Analysis of Indoor Massive MIMO Channels at 6 GHz Using SAGE

In this paper, we present a measurement campaign of indoor massive MIMO channels by using a wideband channel sounder and a 64-element virtual linear array. The measurements are conducted at 6 GHz, with a bandwidth of 200 MHz. Both the light-of-sight (LOS) and obstructed-line-of-sight (OLOS) scenarios are considered in measurements. The Frequency Domain Space-Alternating Generalized Expectation maximization algorithm (FD-SAGE) is used to determine the channel characteristics in angular domain. In order to validate the obtained FD-SAGE estimates, the power azimuth spectrum (PAS) has been calculated using the Bartlett Beamformer (BBF). The non-stationarity of radio channels, which is reflected by power of the estimated MPCs, azimuth of departure (AOD), and azimuth of arrival (AOA), is discussed. The direction spread is estimated, which reflects spatial dispersion of wireless channel. It is also found that the diffuse scattering components at 6 GHz in OLOS scenario is richer than that in LOS scenario, and the magnitude of the diffuse scattering components are comparable with the specular components. It is suggested that the non-stationarity of the diffuse scattering components should not be neglected in the indoor massive MIMO channel modeling at the frequency bands below 6 GHz.

A research on SAGE algorithm based on massive MIMO channel measurements

This paper investigates the channel parameter estimation algorithm by using the Space-Alternating Generalized Expectation-Maximization (SAGE). Two strategies, i.e., parallel interference cancelation (PIC) and serial interference cancelation (SIC), are implemented and validated in the simulated SV channel model. Based on the massive MIMO channel measurement results in an indoor hall scenario in the 6 GHz band, the channel parameters are jointly extracted. The similar results are obtained by using the SIC and the PIC, but the SIC demonstrates lower computational complexity. Finally, the performance of the SIC SAGE is improved by introducing a threshold when estimating the multipath delays. The proposed SAGE algorithm should be useful for extracting the channel parameters with higher accuracy.

Path loss characteristics for vehicle-to-infrastructure channel in urban and suburban scenarios at 5.9 GHz

This paper investigates path loss characteristics for Vehicle-to-Infrastructure (V2I) communications in urban and suburban scenarios. We carried out measurements at 5.9 GHz. The transmitter antenna heights are 3.5 m and 1.5 m and the position of the receiver antenna is on the roof of the car. Then we analyzed the path loss characteristics by using the measured data. The results are helpful for the V2I communications system design.

Characterization of indoor massive MIMO channel at 11 GHz

In this paper, we present a measurement campaign of indoor massive MIMO channels by using a wideband channel sounder and a rectangular large-scale array with 256 elements. The measurements are conducted at 11 GHz, with a bandwidth of 200 MHz, and the light-of-sight (LOS) scenario is considered. Some basic channel parameters, including shadow fading, delay spread and coherence bandwidth, are extracted. The non-stationarity of radio channels, which is reflected by the variations of delay spread and coherence bandwidth over different array locations, is discussed. The results would be useful for the channel modeling of massive MIMO at the frequency bands beyond 6 GHz in the indoor environments.

Empirical evaluation of indoor multi-user MIMO channels with linear and planar large antenna arrays

Channel measurements of large-scale multiuser multiple-input multiple-output (MU-MIMO) radio propagation channels are presented. In the setup, three users with patch antennas communicate simultaneously with a base station (BS) equipped with a large antenna array in an indoor environment. Both a uniform linear array (ULA) and a uniform planar array (UPA) are used, and their relative ability to separate MU-MIMO signals is examined. At the mobile station (MS) side, the effect of inter-user spacing (i.e., the spacing between different users) is investigated. This evaluation is done by means of the correlation matrix distance metric (between each pair of users) and the singular value spread of the system. Our investigation shows that the users can be spatially separated in a large antenna array system in line-of-sight propagation conditions even when they are located close to each other. Furthermore, the users tend to be more separable when a ULA is adopted, compared to using a UPA. Finally, we also confirm that larger user separation distance results in increased channel orthogonality by measurements.

Multi-User Channels With Large-Scale Antenna Arrays in a Subway Environment: Characterization and Modeling

This paper presents results from an extensive measurement campaign of multi-user channels with large-scale antenna arrays (MULAs) in a subway station environment. Based on the measurements, two spatial separation metrics are characterized. The variance of user separability and temporal behavior are investigated. Furthermore, a MULA channel model with transmit correlation is proposed, which ensures that the MULA channels can be generated with a given degree of compatibility. The proposed MULA model is validated by comparing the distributions of the sum rate capacity and the condition number to measurements.