Science Frontier

HUST Prof. Haifan Yin and his team breaks the Curse of Mobility in 5G

Jul 10, 2020

Prof. Haifan Yin and Prof. Yingzhuang Liu of the School of Electronic Information and Communications, Prof. Haiquan Wang of Hangzhou Dianzi University, and Prof. David Gesbert of EURECOM have jointly published their latest findings in IEEE Journal on Selected Areas in Communications (IEEE JSAC, Impact Factor of 9.3), the top-notch journal in the field of communications. The paper, entitled “Addressing the Curse of Mobility in Massive MIMO with Prony-based Angular-Delay Domain Channel Predictions”, breaks the Curse of Mobility in 5G Massive MIMO.

Rooted in large number of antennas at the base station, Massive MIMO is a key technology of 5G for its capability of increasing the performance (spectral efficiency and energy efficiency) of mobile communication systems by at least a factor of 10. However, researchers from industry have identified a severe challenge in field testing and deployment of 5G networks, which is the Curse of Mobility. Previous theoretical papers indicated that Massive MIMO can improve system performance on the order of 10 times. But field tests show that it happens only when the mobile users are moving at a low speed (3km/h for example) or stay stationary. Once the moving speeds of users grow to 30km/h or higher, the performance of 5G system drops 50% or even 80%, as discovered by the testing teams. This problem mainly stems from the fact that the wireless Channel State Information (CSI) is more difficult to acquire in a mobile environment and it ages very quickly. Hence, the mobility problem has become the greatest barrier hindering high-speed data transmission of Massive MIMO. Over several years, telecommunication industry has invested a huge amount of research efforts to solve the problem but to no avail, and thus the problem stands as the "most formidable challenge" confronting the deployment of 5G networks.

Haifan Yin and his team have conducted in-depth research on this problem for a long time, and a new breakthrough have recently been achieved. By exploiting the angle-delay-Doppler structure of the wireless channel and the high angle-delay resolution of Massive MIMO in 5G, the team has invented a Prony spectral analysis-based channel prediction technology, featuring high precision and low complexity. This technology is able to predict future channel variations using very few outdated channel samples, so as to avoid the CSI aging issue as the user moves. The team has proved theoretically that with an increasing number of antennas in Massive MIMO, the channel prediction error of this technology converges to zero for any user moving speed. Taking the 32/64 antenna configuration in current 5G commercial system as an example, the system performance reaches the stationary scenarios, although users are in fact moving at 60km/h, as is shown by evaluation results under an industrial performance validation platform.

Haifan Yin has long been engaged in the research on Massive MIMO, a key technology of 5G. In 2013, he published a paper entitled "A Coordinated Approach to Channel Estimation in Large-Scale Multiple-Antenna Systems " in IEEE JSAC. The findings therein greatly alleviated the pilot contamination problem, which was believed to be the bottleneck of Massive MIMO. Having been cited for more than 1000 times (Google Scholar), the paper gained acclaims from several top international scholars of this field, including Professor Thomas Marzetta, the inventor of Massive MIMO and member of the National Academy of Engineering of the US. After pilot contamination has been essentially solved, the mobility problem arises as the biggest challenge confronting Massive MIMO. After more than one year of research efforts, Haifan Yin and his team once again achieved technological breakthrough, which greatly alleviated the Curse of Mobility in 5G deployment.

Publication Link:

Source: School of Electronic Information and Communications

Address: Luoyu Road 1037, Wuhan, China
Tel: +86 27 87542457    Email: (Admission Office)

©2017 Huazhong University of Science and Technology