Exit Menu

Beamforming using Artificial Intelligence for 6G Networks


Project description

Massive multiple-input multiple-output (MIMO) systems have shown promising features to significantly increase spectral efficiency of cellular communication systems.

Future wireless communications such as 6G requires implementation of such systems in extremely high frequency (EHF) range up to 300 GHz for various reasons such as shortage of available spectrum and low attenuation, within 1-2dB/km. Deployment of massive MIMO systems in EHF band, i.e., millimetre-wave (mmWave) will result in spectral efficiency as well as increased available bandwidth.

Nevertheless the severe attenuation of mmWave signals is a dominant obstacle, and is generally compensated via beamforming techniques that employ the benefit of large antenna arrays in massive MIMO structures often embedded in a tiny dimension in mmWave frequencies. Analog beamforming has shown extensive advantages compared to its traditional counterpart, digital beamforming. Dominant techniques proposed so far for analog beamforming are often practically challenging. Down-to-earth confrontations are namely the necessity of possessing excellent channel state information (CSI) by the base station (BS) or search complexity. Reduction in search complexity can be achieved by performing high number of iterations between user and BS to exchange information; this leads to inevitable practical overhead.

In our research, we plan to propose intelligent beamforming (IB) schemes in contrast. IB is achievable by deployment of artificial intelligence (AI) techniques through design structures that can margin novel context-awareness beamforming to drive 6G. Such implementation will aggregate the advantages of combined massive MIMO and mmWave systems, specifically when designed in upper band EHF such as 200-300GHz, and at the same time intelligently diminishes the complexity and overhead of existing schemes to make it a resourceful option for engineering 6G.

The next generation Communication system will benefit from the results and this project will open the door for more research in future applications for the connectivity between users with high data rate.

One of the benefits of mmWave characteristic is small wavelength, for example in 240GHz frequency the half wavelength is = 0.625mm, so a 10x10 antenna array would be in about a size of portcullis of penny.