5G System-Level Simulation results interpretation: BLER = 1 vs non-null goodput/throughput

Hello,

I am simulating one UE at 10 km from gNB using 5G System-Level Simulator. The BW = 20 MHz, and UE EIRP = 23 dB and channel is CDL-A (NLOS) / CDL-D (LOS), with delay spread of 260 ns.

I am getting some throughput and goodput even when BLER is 1 (see the results below).

Is this correct? either yes or no can you please explain why? Is this related to HARQ?

Also given the 3GPP target of 10% BLER, should i still consider performance values for BLERs > 10%.

"Peak UL Throughput: 131.89 Mbps. Achieved Cell UL Throughput: 65.53 Mbps

Achieved UL Throughput for each UE: [65.53]

Achieved Cell UL Goodput: 16.77 Mbps

Achieved UL Goodput for each UE: [16.77]

Peak UL spectral efficiency: 6.59 bits/s/Hz. Achieved UL spectral efficiency for cell: 0.84 bits/s/Hz

Block error rate for each UE in the uplink direction: [1]."

Thanks in advance.

Answer 1

[Naren]

Hello Houcine,

Yes, it is correct. Even with a block error rate (BLER) of 1, some throughput and goodput are still achievable. This is because 5G networks use the Hybrid Automatic Repeat Request (HARQ) technology. Retransmitting incorrect packets is made possible by HARQ, raising the likelihood of a successful transmission, and lowering the effect of mistakes on overall performance.

Considering the 3GPP aim of 10% BLER, it is nevertheless critical to consider performance values for BLERs exceeding 10% since they can significantly affect network performance, especially in congested or heavily loaded networks. It is also crucial to keep in mind that, even with HARQ retransmissions, the feasible throughput and goodput values may drop as the BLER rises.

You can refer to 3GPP documentation for further information.