Massive public events requires a special type of network intelligence.
Variance is the nightmare of any process. Sudden oscillations in demand are, for example, at the heart of traffic jams, hospital queues and electrical blackouts. Unsurprisingly, then, nothing is more conducive to congested mobile telecommunications networks than public events, when thousands of people show up in the same place while still eager to remain 100% ‘online’.
After observing the cells serving a sporting event, we noticed that the top 3 most used applications were Instagram, Facebook and YouTube, in that order. More interestingly, the split between content upload versus content download was tilted towards downloads. In other words, subscribers were not so much uploading their pictures or videos as much as they were, instead, just browsing their feeds.
“Nowadays, people can’t let go of their phones even during the live experiences that they paid to enjoy.”
Large-scale events require the deployment of additional infrastructure to meet the concentrated data demands of a densely packed area. The only way to achieve that effectively is for network planners to leverage insights into the utilisation and performance of the network, both historically and in real-time.
With that, they can drive more accurate network interventions – and learn from them. They can determine, for example, whether the change in cell configuration and cell deployment for the same event across two distinct years has improved subscriber experiences. Or, then, if the intervention on the 3rd day of a 5-day event produced the desired results.
Real-time observations should also steer better ‘command centre’ decisions, by quickly identifying areas with degrading experience and drilling down to cell level for troubleshooting. The immediate tracking of what is happening in each point of interest/group of cells should define which cells to offload subscribers to, or then trigger other strategies (such as adjusting antenna angles).
Network planners must also understand the changing trends in in-event subscriber consumption and identify the main contributors to uplink and downlink traffic. Based on our experience, temporary spikes in uplink traffic during “instagramable moments” are often accompanied by significantly degraded subscriber experiences.
By equipping themselves with the right network analytics intelligence, network planners may come to fear less the swings in users and throughput that public events trigger. And mobile operators may have the opportunity to stick out positively in their markets, turning what could have been a troublesome collective experience into a seamless, event-worthy one.