Enterprises willing to pay for guaranteed latency

Spirent Communications  has released a benchmarking analysis study conducted with STL Partners based on real-world mobile edge network testing and 150 interviews. The "Cutting Through the Edge Computing Hype - MEC Latency Expectations vs. Reality" report provides critical insights into multi-access edge computing (MEC) performance measurements that matter most, what can be achieved via today’s networks, and use cases with the greatest potential for near-term monetization. 

“In 2021, Spirent saw service provider 5G engagements increase by more than 50% year-on-year, as customers raced to make 5G a reality,” said Spirent’s Head of Market Strategy, Steve Douglas. “Market competition intensified, and the impact of the pandemic accelerated business automation and monetization plans. As a result, service providers established early partnerships with public cloud providers with heightened focus on edge-delivered low latency services, which is where our report is focused.”

With 5G comes the desire for greater responsiveness that is as close to real-time as possible, as Spirent Sr. Director of Mobile Service Strategy, Rich McNally explains: “We work with the world’s leading service providers and the expectation of near real-time latency has generated keen interest in MEC network architectures. Pushing cloud computing capabilities to the edge of the network is seen as a way to provide enterprises with improved service reliability via consistent, optimal latency and less jitter.

“Our study shows that the industry is off to a solid start, but there is still work to do. What we’re seeing through our 5G benchmarking service engagements is that the latency of real-world MEC services can fluctuate significantly - by time and across regions – with a lack of symmetry between the uplink and downlink,” said McNally.

“Ultimately, latency must be managed holistically, and end-to-end, to achieve reliable and desired customer experiences, and to meet service-level agreements (SLAs).”

The report is based on Spirent’s real-world edge network testing engagements and more than 150 interviews with prospective edge customers conducted by telco industry analysts STL Partners. Key findings include:

•       Latency Consistency – Enterprise users overwhelmingly prioritize latency consistency to support the edge applications they plan to deploy. 56% would be willing to pay for an SLA with guaranteed latency that never exceeds a predefined window. The vast majority (66%) needed latency of 50ms or less. The window needed most, at 37%, was 20-50ms of latency.

•       The Disconnect – Edge application demand and supply sides are disconnected in terms of what latency will be required, when it will be required, and for which use cases. There is a lack of data-based understanding of actual latency numbers/ranges required. There is also misalignment when the network and 3GPP releases can actually support the requirements.

•       Short-Term Edge Opportunities – Gaming, video analytics and AR/VR dominate the short-term edge opportunity. They have a strong requirement for low latency, and will continue to evolve rapidly with unique requirements, so are out of sync with current reality – at least until consistently low latency is provided through 5G edge solutions. Cloud gaming is directly impacted by latency, latency fluctuations and symmetrical bandwidth, but once the edge can provide those capabilities, video gaming will use them.

•       MEC Latency by Region – The benchmarking results reveal a wide variation in latency among regions, not only between cloud and MEC implementations, but also between different markets within the same region. Currently, guaranteeing a national level SLA would be problematic. It would also be challenging for an app developer to design a solution intended for global utility if the latency window varies wildly from city to city.

Recommendations for Improving MEC Latency

• Establish the right test regime. Improvement starts with measurement. Application testing profiles should take into account data volume to emulate required throughput, packet/frame size to emulate segmentation, and packet velocity to assure holistic results. Testing should be performed in all target markets to measure consistency of performance in all customer locales.

• Realize the benefits of 5G standalone upgrades. Early upgrades to 5G standalone (SA) have demonstrated improvements in reducing network latency by over 20%, with more than 450 devices and form factors already supporting 5G SA.

• Private MEC vs. public MEC for industry. For enterprises and industries, the implementation of a private 5G MEC on the premises offers the potential for further latency reduction due to the proximity for localized processing. Local breakout can be provided to access the public cloud and MEC, while the private MEC can be optimized for latency-sensitive applications where over-the-air and transport network impact performance.

• Implement 3GPP release upgrades. 5G standards development has focused on a continued stream of enhancements for reliable low latency.