Investigating 4G and 5G for Live Football Streaming
The work included extended 4G monitoring, assessment of streaming bandwidth requirements, investigation of fixed-line alternatives and consideration of a wider shared-connectivity solution for the sports site.
Live video streaming places a particular demand on an internet connection: consistent upload capacity is more important than headline download speed alone. The club's existing connection produced a reported speed-test result of approximately 6 Mbps, below the streaming provider's initial stated upload requirement of 15 Mbps.
Kent-ITS investigated the available alternatives and collected approximately 2,335 records during an extended 4G monitoring period. Preliminary analysis of successful throughput tests indicated typical download speeds of around 190 Mbps, upload speeds of around 31 Mbps and latency of approximately 25 ms.
These results demonstrated that the tested 4G connection could deliver substantially more bandwidth than the initial clubhouse measurement suggested was available from the existing service. However, upload performance varied, and some observations fell below the subsequently confirmed minimum streaming requirement of 10 Mbps.
The investigation also considered 5G, satellite broadband and dedicated fixed-line connectivity. The findings provided an evidence base for further planning, but did not constitute a live-streaming trial or proof of a completed 5G or satellite installation.
The Challenge
The club wanted to stream live football matches but needed to establish whether its internet connection could reliably support the service.
The initial clubhouse speed test indicated approximately 6 Mbps. The streaming provider initially specified 15 Mbps upload, before confirming that its variable-bitrate system could adapt to a minimum of 10 Mbps. A connection that occasionally reached the required speed would not necessarily be suitable: the available upload capacity needed to be assessed over time and under changing network conditions.
There were several additional constraints:
- Fixed-line availability: Full-fibre broadband was not available at the site during the investigation.
- Cost: A dedicated leased-line option was quoted at a minimum of £500 per month plus VAT on a three-year term, making it a significant recurring expense.
- Mobile performance: 4G and 5G performance can vary with radio conditions, network loading and other factors. Headline speed tests alone could not establish sustained upload suitability.
- Satellite connectivity: Starlink was considered as another potential option, but no installation or suitable means of collecting comparable real-world performance data was available during the investigation.
- Future requirements: Connectivity might eventually serve more than the streaming operation, creating an opportunity to consider a shared network across the wider site.
The task was therefore to gather useful evidence, understand the practical limitations and compare potential approaches without assuming that any untested option would meet the requirements.
The Solution
1. Establishing the bandwidth requirement
The streaming provider's requirements were reviewed to distinguish the initial 15 Mbps upload recommendation from the later-confirmed 10 Mbps minimum for variable-bitrate streaming.
This established a measurable target for the investigation. It also highlighted the importance of retaining sufficient headroom above the minimum, rather than treating a single successful speed test as proof of suitability.
2. Extended 4G monitoring
A 4G connection was monitored over an extended period, from 25 June to 16 September 2026. Approximately 2,335 records were collected using a MikroTik monitoring setup with an LTE modem and an existing business mobile-data connection.
The collected data provided a broader view of performance than an isolated speed test. Download throughput, upload throughput and latency were considered alongside failed or incomplete throughput tests and changes in observed radio performance.
The results also demonstrated why individual measurements require context. Some failed tests recorded zero throughput despite latency and packet-loss readings that did not, by themselves, indicate a complete connection outage. Such records should not automatically be classified as confirmed service outages.
3. Reviewing the measured performance
Preliminary analysis of successful throughput tests indicated:
- Typical download throughput of approximately 190 Mbps.
- Typical upload throughput of approximately 31 Mbps.
- Latency of approximately 25 ms.
- Upload observations ranging from roughly 5 Mbps to 53 Mbps.
Some download results exceeded 250 Mbps, while upload results were not consistently above the 10 Mbps minimum. A period of radio-performance degradation was also observed in early August before performance recovered.
These findings suggested that 4G had considerable bandwidth potential at the test location, but that upload variability remained a relevant consideration for live streaming. The figures are preliminary and should be checked against the original dataset before being published as final statistical results.
4. Assessing fixed-line and 5G alternatives
Fixed-line availability was investigated, but full-fibre service could not be obtained at the site during the project. A dedicated leased line was considered, although its quoted recurring cost was substantial for the immediate requirement.
A dedicated 5G installation was also considered as a potential alternative. The proposed approach involved suitable outdoor 5G equipment, an appropriate business mobile service, routing and firewall facilities, switching and structured cabling.
This remained a proposed solution rather than an installed and tested system. The 4G results supported further investigation of mobile connectivity, but they could not be used to claim that a 5G installation would achieve the same or better performance.
5. Considering satellite broadband
Starlink was examined as a possible alternative to terrestrial broadband. However, no service was installed and no suitable means of carrying out a comparable real-world performance assessment was available during the investigation.
No formal Starlink quotation was obtained, so its final installation cost, recurring charges and suitability for the site remained unverified. It was therefore not possible to make an evidence-based comparison of actual Starlink performance against the monitored 4G connection.
Satellite broadband remained an option for further evaluation rather than a tested or conclusively rejected solution.
6. Exploring a wider shared-connectivity design
The investigation also considered whether a future 5G connection could provide connectivity for more than the streaming operation.
One concept was to distribute connectivity between buildings using point-to-point wireless links, with separate network segments, routing and firewall controls for different users or services. Depending on requirements, this could potentially support additional facilities such as Wi-Fi, CCTV or IP telephony.
This was a proposed architectural option only. It was not implemented, and the cost and practical benefits of a wider site network had not been established.
Project Outcomes
The investigation produced a useful body of evidence for assessing internet connectivity for live football streaming.
The principal outcomes were:
- Measured 4G performance: Extended monitoring provided a more representative view of mobile broadband performance than a single speed test.
- Bandwidth potential identified: Preliminary results showed substantially more upload capacity than the initial clubhouse speed-test result, although performance varied over time.
- A clear minimum requirement established: The confirmed 10 Mbps streaming threshold provided a practical benchmark against which the measured upload results could be considered.
- Fixed-line limitations documented: Full-fibre availability and the cost of a dedicated leased line were investigated, establishing important constraints on the available choices.
- Further mobile investigation identified: 5G was retained as a potential route forward, subject to a separate installation and performance assessment.
- Satellite uncertainty documented: Starlink could not be validated because no installation or comparable real-world testing was undertaken.
- Wider networking possibilities identified: A shared connectivity design was considered for potential future use across the site, without presenting it as an implemented solution.
The key outcome was an evidence-led assessment of the available options, including their limitations and the work still required to validate them.
The investigation did not establish a guaranteed streaming service, nor did it confirm the performance of an installed 5G or satellite solution. Instead, it provided a technical basis for deciding what should be tested next and for avoiding investment decisions based solely on headline speeds or unverified assumptions.