This resource note presents empirical measurements of a Quectel 5200 modem utilizing IoT slicing (SST3) in two distinct network conditions. The first condition represents an unloaded network with minimal traffic. The second condition features a heavily loaded network on a different slice, SST1.

The primary objective of this test is to gain a deeper understanding of how network slicing performs in an environment where other network slices are experiencing overload.This resource note details the methodology and presents empirical measurements from a crucial performance evaluation of 5G network slicing, specifically utilizing open hardware and open software components. The test environment focused on assessing the resilience and isolation properties of a designated IoT slice (Service/Slice Type – SST3) under varying network load conditions.

1. Hardware and Software Configuration

The core hardware component under investigation was a Quectel 5200 modem, selected for its commercial availability and support for 5G slicing features. The underlying network infrastructure employed an open-source 5G core and RAN (Radio Access Network) solution, emphasizing the objective of validating performance in an open, non-proprietary ecosystem.

2. Testing Scenarios and Objectives

The primary objective of this testing campaign was to gain a deeper, quantitative understanding of the performance and isolation capabilities of 5G network slicing, particularly how a low-priority slice is affected when a high-priority, different slice experiences severe congestion. To achieve this, measurements were taken for the SST3 slice under two distinct, controlled network conditions:

2.1 Condition 1: Unloaded Baseline (Minimal Traffic):

• • In this scenario, the network as a whole, including both SST3 and all other available slices (SST1), was operating with minimal background traffic.
  • This condition established the performance baseline for the Quectel 5200 modem on the SST3 slice, reflecting its optimal throughput, latency, and reliability metrics without external interference or resource contention.

2.2 Condition 2: Heavily Loaded Competitor Slice (SST1 Overload):

• • This scenario introduced a controlled, heavy traffic load specifically onto a different slice, SST1, which was provisioned with higher priority than the target SST3 slice. SST1 is typically used for enhanced Mobile Broadband (eMBB) or other high-throughput applications.
  • The traffic injection was designed to simulate a real-world overload or congestion event on the SST1 slice.
  • The critical focus of this test was to measure the resulting performance degradation (or lack thereof) on the low-priority SST3 slice. The data collected will quantify the degree of slice isolation achieved by the 5G core and RAN, indicating whether the severe resource contention on SST1 bleeds over and negatively impacts the dedicated resources of the SST3 slice.

3. Expected Outcome

The testing aimed to gather empirical data on the uplink and downlink throughput of the device under test (DUT). This data was collected using the 5G Channel 78 across various transmission bandwidths, which are illustrated in the figure below. A key design objective is to ensure that the performance of the DUT, when utilizing the SST3 slice, remains unaffected by traffic on other slices.