Introduction

Assessing mobile broadband for live video streaming requires more than checking a headline speed. Download performance may be impressive, but the connection must also sustain sufficient upload capacity to transmit the video stream. Changes in radio conditions can affect throughput and latency, making measurements over time more useful than an isolated speed test.

This investigation examined 4G performance between 24 June and 15 October 2026, using individual throughput-test results alongside telemetry recording radio conditions and connection performance.

The streaming requirement

The initial clubhouse speed test reported approximately 6 Mbps. The streaming service initially specified an upload requirement of 15 Mbps, subsequently confirming that its variable-bitrate system could adapt to a minimum of 10 Mbps.

These thresholds provided a practical basis for assessing the measurements. Results below 10 Mbps were particularly significant because they fell below the confirmed minimum, while results between 10 and 15 Mbps remained below the original requirement.

The monitoring approach

The monitoring setup used a MikroTik platform with a Quectel EG18-EA LTE modem. The archive contained 3,186 individual JSON files, of which 2,867 contained completed throughput results. Additional telemetry recorded radio measurements, including RSRP, RSRQ and SINR, alongside throughput and latency.

The distinction between a completed test and a failed or missing result is important. The remaining files contained errors or no data, so they were not treated automatically as evidence of a network outage.

Overall performance

Across the 2,867 completed throughput results, median download speed was 191.46 Mbps and median upload speed was 32.04 Mbps. Median latency was 25.36 ms.

These results demonstrate substantial bandwidth potential under the conditions observed during the monitoring period. However, the overall figures do not tell the whole story.

In 118 completed results, approximately 4.1% of the total, upload throughput was below 10 Mbps. A total of 273 results, approximately 9.5%, were below 15 Mbps. The connection therefore delivered ample upload capacity in most completed tests, but it did not consistently remain above either threshold.

The significance of radio conditions

The accompanying radio telemetry helps explain the most pronounced changes in performance. Under the more usual conditions recorded during the investigation, RSRP was commonly around −81 to −82 dBm. During periods of significant degradation, the recorded signal level fell considerably and throughput deteriorated at the same time.

On 3 August, evening measurements recorded download speeds of approximately 13–23 Mbps and upload speeds of around 3–4 Mbps. RSRP fell to approximately −116 to −117 dBm, while latency increased.

A second pronounced episode occurred on 24 September. Median download throughput fell to approximately 12.73 Mbps, median upload to 7.91 Mbps and median latency increased to around 44 ms.

The clearest sustained example occurred on 13 October. Median download speed fell to 23.91 Mbps, median upload to 5.67 Mbps and median latency rose to 55.26 ms. Radio telemetry recorded RSRP around −112 to −115 dBm during much of that day.

These observations show a strong association between degraded radio conditions and reduced connection performance. The measurements identify the timing and scale of the deterioration, although they do not establish its underlying cause.

Recovery and comparison

The deterioration on 13 October was not permanent. Later that evening, throughput and radio measurements returned towards the levels seen during normal operation. Results from 14 and 15 October confirmed that this recovery continued, with median download performance around 191 Mbps and upload around 35 Mbps across successful telemetry measurements. Median RSRP was again approximately −81 dBm.

This comparison is particularly useful because it demonstrates the difference between the connection's performance under normal radio conditions and during a period of severe degradation. A single speed test taken during either period would have provided a very different impression of the service.

Conclusion

The monitoring established that the tested 4G connection could deliver substantial bandwidth, with median upload performance comfortably above the streaming service's confirmed 10 Mbps minimum. It also showed that this capability was not consistent under all observed conditions.

The most important finding is the relationship between radio deterioration and the loss of usable upload capacity. During the more severe episodes, upload performance fell well below the streaming threshold before recovering as radio conditions improved.

For applications such as live video streaming, mobile broadband should therefore be assessed using extended throughput and radio monitoring, with particular attention to the lower end of the performance range. Headline speeds and overall medians describe the connection's potential, but the periods of degradation reveal the limitations that matter when a continuous upload is required.