1 of 1

Introduction

Conclusions

Low complexity beamforming designs have been proposed for 2-dimensional terrestrial mmWave communication, UAV-assisted mmWave communication, and aerial RIS-assisted mmWave communication
A variable time frame structure with TTI of 1 OFDM symbol along with an IA procedure for sectored-cell model has been proposed to improve the spectrum utilization in case of non-homogeneously distributed UEs
Via simulation results, it has been shown that in sectored-cell system model, using narrowest beam to serve UEs is not optimal in single RF chain assisted as well as in a multi-RF chain assisted mmWave communication system
Localization error does not affect the optimal sectorization scheme (equivalent to optimal beamwidth)
To leverage beam squint, a RF-SC-BB based joint beamforming and OFDM subcarrier allocation has been proposed that provides improved spectrum efficiency
Further, it has been shown that in UAV-assisted mmWave communication system the clustering based approach outperforms sectoring based approach
For an aerial RIS-assisted mmWave communication, a reduced complexity AoA based RIS beamforming and UE beamforming design has been proposed along with an AoA estimator employing only one RF chain

Bharti School of Telecommunication Technology and Management

Industrial Significance

The proposed sectored-cell methodology is useful for deployment of standalone wideband mmWave communication networks when the number of devices to be connected is significantly higher than the number of available RF chains
The proposed beamforming methodologies, for both terrestrial and aerial mmWave communications, offer enhanced spectral efficiency and energy efficiency using restricted number of RF chains that lowers the deployment cost and hardware complexities, desirable in 5G and beyond communication networks

Beamforming at mmWaves for

5G and Beyond Communication Networks

Nancy Varshney and Swades De*

Industry Day Theme #Communication Technologies

References

[1] N. Varshney and S. De, “Optimum Downlink Beamwidth Estimation in mmWave Communications, " IEEE Transactions on Communications, vol. 69, no. 1, pp. 544-557, Jan. 2021.

[2] N. Varshney and S. De, “Multi-RF Beamforming-Based Cellular Communication Over Wideband mmWaves," IEEE Transactions on Communications, vol. 70, no. 4, pp. 2772-2787, April 2022.

[3] N. Varshney and S. De, “RF Beamforming and Subcarrier Allocation Using Beam Squint in mmWave Systems," IEEE Wireless Communications Letters, vol. 11, no. 4, pp. 678-682, April 2022.

[4] N. Varshney and S. De, “Design Optimization for UAV Aided Sustainable 3D Wireless Communication at mmWaves, " IEEE Transactions on Vehicular Technology, Early Access, 2022.

Acknowledgement

This work was supported by the DST under Grant INT/Korea/P-46, the DoT under Grant 4-23/5Gtest-bed/2017-NT, and the SERB under the Abdul Kalam Technology Innovation National Fellowship.

Figure 5: Proposed versus existing scheme of subcarrier allocation, RF precoding, and baseband precoding in mmWaves.

Figure 10: Illustration of average long-run UE rate as a function of number of sectors and number of RF chains for M=100 UEs.

Figure 16: Improvement in (a) average of long-run UE rate and (b) geometric mean of long-run UE rate achieved using variable frame with TTI=1 OFDM over variable frame but TTI=14 OFDM symbols and over equal frame time for sectors. (M=100)

Figure 9: Normalized interference experienced by a UE with respect to variable half power beamwidth (HPBW).

(a) (b)

Figure 12: Collision probability versus number of new UEs arriving during a sector sojourn time.

Figure 13: Improvement in IAD using proposed initial access procedure over the 3GPP NR.

Figure 14: Saving in RBs using proposed initial access procedure over the 3GPP NR.

Figure 22:: Comparison of the spectral efficiency achieved with the proposed AoA based method over the existing schemes in literature of an aerial RIS-assisted mmWave communication system.

Figure 23: Comparison of the computational complexity of the parameter estimation plus RIS weight optimization of the proposed AoA based approach to that of estimated CSI based approach in an aerial RIS-assisted mmWave communication system.

Figure 17: Aerial RIS-assisted mmWave communication system to serve a UE located on the ground in the shadowed region. (a) Shows using k-means clustering to group and serve UEs. (b) Shows using sectorization method to group and serve UEs.

Figure 3: (a) Pictorial representation of true position X and estimated position X’ of a UE along with the actual beam coverage region considering practical radiation pattern. (b) Demonstration of effective AoA of the UE; Ψ is the true AoA and Ψe is the estimated AoA with respect to sector 2, whereas Ψa is the true AoA of UE with respect to sector 1.

Figure 11: (a) 3GPP NR initial access (IA) procedure; (b) Proposed IA procedure.

Figure 15: Proposed variable time frame structure for single sector in sectored cell model.

Figure 1: (a) and (b) depicts 2-dimensional round-robin scheduling of S=12 sectors over two consecutive epochs with one RF chain (i.e., one analog beam).

Figure 8: (a) and (b) depicts 2-dimensional round-robin scheduling of S=12 sectors over two consecutive epochs with number of RF chains (or beams) NRF= 4. Sectors are represented as {S1,..., S12} and beams are represented as {B1,..., B4}. Here, number of sector served by beam B1 = {S1, S2, S3}. (c) Shows steering directions {R1, ..., R8} of a UE within beam B1.

Figure 18: Plot of gain in geometric mean rate achieved by using k-means clustering over sectoring approach at mmWaves, given power limit power of communication power = 5 W.

Figure 19: Illustration of optimal number of beams as a function of available power for communication module at UAV to serve non-homogeneously distributed UEs on ground at mmWave range

Figure 20: Comparison of achievable geometric mean rate at sub-6 GHz and mmWaves at the UAV as a function of power budget and illustration of backhaul power consumption at sub-6 GHz.

Figure 2: (a) Average long run UE rate as a function number of sectors (inversely proportional to sector HPBW).

(b) Energy efficiency of the system as a function number of sectors.

(c) Trade-off between average long-run UE rate and energy efficiency in estimating optimal sectorization scheme (optimal beamwidth).

Result

Figure 4: Average long-run UE rate versus total number of sectors plot for (a) 1000 UEs uniformly distributed in the cell, and (b) 500 UEs non-uniformly distributed in the cell with probability distribution vector [0.9, 0.05, 0.03, 0.01, 0.008, 0.002].

Figure 7: Gain in spectral efficiency using RF-SC-BB scheme over SC-RF-BB scheme at 28 GHz and 5 GHz.

Result

(a) (b)

Advantages of mmWaves

Research flowchart