LTE vs 5G NR

[henry3499]

Practical coverage and capacity simulation comparison between LTE 1800 MHz and 5G NR 3500 MHz using Infovista Planet 7.2 is demonstrated in this article.
The comparison cover the following LTE and 5G NR network performances:

The simulation inputs comparison is summarized in below points:
LTE
· UE Specification
o Max TX Power: 23 dBm[/COLOR]
o Antenna directivity: Omnidirectional[/COLOR]
o Antenna gain: 0
· RBS Specification
o Freq Band: L1800 MHz
o Bandwidth: 10 MHz
o TX/RX: 2/2
o Loading: 50%
NR
· UE Specification
o Max TX Power: 23 dBm
o Antenna directivity: Omnidirectional
o Antenna gain: 0
· RBS Specification
o Freq Band: NR3500 MHz
o Bandwidth: 100 MHz
o TX/RX: 64/64
o Loading: 50%

The power inputs is managed in a way to have similar LTE and NR Reference signal (RS) Power

Propagation Models
The right propagation model with an optimum settings is the key for valid simulation and comparison. Infovista Planet 7.2 planning tool is used for the simulation and P3M propagation model with buildings information is the main used model for quality and capacity plots. I intentionally created additional coverage plots using Planet 2D General Model (PGM) and Atoll Aster 3D model. The statistics proved that 2D models provide different coverage statistic compare to P3M and also by using different other planning tool model like Atoll Aster 3D , the predictions are slightly different.

It is important to refer to a drive measurement to validate the final propagation model to be used.

P3M Propagation Model
Below figure is showing P3M Model (3D) settings used for LTE 1800 and NR 3500.
The main difference reside in Frequency , outer wall loss and penetration loss slope.

Planet General Model (PGM)
The 2D models like PGM is based on "K" factors and are mainly slope ("K2": horizontal attenuation) and intercept ("K1" : Frequency offset).
The "K2: is frequency independent and depend on clutter type (Dense Urban, Urban,..) and "K1" is frequency dependent. In our case different "K2" is set per clutter type and the same "K2" is set for both bands (LTE1800 and NR3500) but with different "K1".
In case of PGM model, it is possible to use calibrated PGM 1800 MHz band and conclude the frequency shift to be added for "K1" using following formula:
K1(3500 MHz) = 33.9 * Log (3500/1800) = 9.8 dB. Which means K1(3500) = K1(1800)+ 9.8.
Below figure shows an example of PGM (2D) configuration settings in Dense Urban environment:

Atoll Aster 3D model
Aster model is ray-tracing model with high accuracy and it is good model to use for 5G design. It provide flexibility in providing various settings for different propagation class.
Below figure is showing Aster Overall Settings:

In next, will share the simulation output for different network performance.
The left side figures correspond to LTE and the right one to 5G NR!Best Server Area - P3M
The plot is showing the coverage area for each cell and each color correspond to a best server cell coverage for LTE and 5G NR network.

SS-RSRP [dBm] - P3M
The above figure shows the coverage comparison of LTE 1800 and NR3500.

The coverage of LTE1800 is showing better signal strength due to low penetration loss and other factors.
SS-RSRP [dBm] - PGM
Using Also PGM model, LTE stats are showing better coverage but the signal is almost attenuated with fixed offset in all the pixel. This is normal since we have just applied an offset in the PGM formula. For sure power difference, antenna gain and other gain plus loss will be considered while doing the coverage calculation. Like if 5G antenna gain is 2 dB higher than LTE, this will reflect in the coverage compensation calculation and not only the frequency offset loss!
Aster 3D model prediction is showing deviated statistic since we have used the Aster default settings but it is possible to adjust the aster 3D model settings in order to match the P3M coverage statistics.
Below figure illustrate the CDF comparison between 4G and 5G for P3M and PGM Model.

Good to notice that PGM model is showing constant delta difference between 4G and 5G which is due to frequency offset (K1) but in case of PGM the delta varies at different signal levels which is due to other factors are considered like the penetration loss inside the buildings and the complexity of 3D model.
SS-RSRQ [dB]
Below figure is showing the SS-RSRQ comparison between LTE and NR:

The SS-RSRQ is better in 5G due to better dominance.
PDSCH C/(I+N) [dB] – P3M

The PDSCH C/(I+N) is showing great improvement in 5G and it is mainly due to AAS antenna (Massive MIMO) which is using a narrow beam that minimize the overlap area with other cells.
PDSCH Maximum Achievable Data Rate [Mbps] – P3M

Due to improvement in PDSCH C/(I+N) and higher 5G bandwidth , the user throughput in 5G is better than the user throughput in LTE network.
PUSCH C/(I+N) [dB] – P3M

The PUSCH C/(I+N) is worst in 5G and this is because of high number of resources element used in up-link compare to LTE.
PUSCH Maximum Achievable Data Rate [Mbps] –P3M[/COLOR]

Since the up-link bandwidth is higher in 5G network, the up-link throughput is better.
[COLOR=rgba(0, 0, 0, 0.7)]PDSCH Maximum achievable spectral Efficiency – P3M
The below figure is showing the spectral efficiency of LTE 1800 and 5G 3500.
The spectral efficiency is define as Throughput / bandwidth and expressed in bits per second, per Hz (bps/Hz).
It is expected that 5G NR spectral efficiency should be better than LTE.

Conclusion
As expected, the 5G NR is showing better network performance in quality and user throughput but worst in coverage compare to 4G network at sub band spectrum (3.5 MHz). The coverage issue will result in low cell radius and therefore more required 5G sites. For accurate comparison between 4G and 5G, it is important to set the right inputs specially for the networks parameters settings and the use of valid propagation model.
Thanks for reading and kindly share your thoughts !

[islam wagih]

Could you share EXCEL tool and PDF please?