5G creates new network optimization challenges even for seasoned operators. Here are four such challenges, and how operators can overcome them.
Network optimization is a fact of life in the telecoms business for fairly obvious reasons. Operators want their networks to run at peak efficiency at all times, not only for the cost savings, but also to make sure that customers have little reason to complain – or worse, churn – because of poor service quality.
Network operators have years of experience in making sure their networks are as efficient as possible. But each generation of cellular technology brings new challenges in network optimization, and this will be especially true for 5G – possibly more so than operators may be expecting.
The anticipation of 5G’s arrival has created very high expectations – not only regarding data speeds and low latency, but also the new kinds of services these capabilities will enable. Consequently, network quality will matter more than ever – operators can’t afford to promote the wonders of 5G to consumers and enterprises alike only for their network performance to fall short.
Meanwhile, experience with 3G and 4G network optimization will only get you so far – 5G brings plenty of new things to the table that collectively create a number of new optimization challenges for operators. Here are four such challenges, and suggested solutions to overcome each of them.
1. Dual-mode networks
The first phase of 5G rollouts involves the Non-Standalone (NSA) 5G standard approved in the middle of 2018 by the 3GPP. NSA 5G is designed to be deployed on top of existing 4G network infrastructure, to include the base stations. In essence, the result is a coordinated 4G/5G network, which creates serious optimization challenges. Put simply, the 4G network also has to carry 5G signaling, which means it has to be optimized to keep up with the performance demands of 5G so customers don’t notice a sudden drop in service quality when their connection drops down from 5G to 4G.
A related challenge is the fact that many NSA 5G networks will be dual-mode, with massive MIMO arrays supporting both 4G and 5G. This means things like frequency, power, azimuth, and downtilt have to be configured for eNodeB or gNodeB to ensure 4G network coverage and 5G performance isn’t compromised.
This could be achieved by using big data analytics to predict 4G network capacity requirements, and by formulating power and parameter solutions based on test and verification data on the live network to coordinate optimization of 4G/5G networks and improve performance.
2. Massive MIMO complexity
Speaking of massive MIMO, this key component of 5G comes with its own set of capacity and optimization difficulties, mainly because it’s far more complex. For 4G antennas, there are only about 200 parameter combinations. With 5G, we’re talking about more than 10,000 parameter combinations. That makes it even more difficult to determine the optimal parameter combination for each base station.
According to Huawei, this is where AI comes in really handy. Huawei explains that it has introduced AI capabilities to its network optimization tool platform, enabling it to notonly predict 5G network coverage based on data from existing 4G base stations and identify potential future 5G network problems in advance, but also to optimize multiple parameter combinations for massive MIMO.
Huawei adds that this isn’t just about running algorithms that tell you how to tweak the network here and there – the company has developed what it calls an “expert + tool platform” model that combines the optimization tool itself with a team of experts participating in the optimization process.
3. Refarming spectrum
Another optimization challenge is spectrum refarming. 5G works best with at least 100 MHz of spectrum, but some operators may not have that much spectrum available. Refarming 4G spectrum is a way around that – especially for NSA 5G – but this could create interference issues between 4G and 5G NR RANs. It also runs the risk of degrading 4G capacity, which in turn impacts the performance of 4G-based services.
Consequently, operators need to fully evaluate how refarming will impact network performance in terms of 4G capacity and signal interference. In the case of interference, that also means understanding which radio (LTE or 5G NR) is experiencing greater interference.
4. Service optimization
A fourth challenge is optimizing the actual new services 5G will support (as opposed to the network itself). This matters because 5G is expected to introduce all kinds of new services that 4G not only has never had to support, but also isn’t designed to support in the first place – particularly low-latency services such as live virtual reality (VR) video streams or live VR eSports games. Unlike 4G, 5G networks are dedicated to experience modeling and network optimization based on low latency. This gets complicated because different 5G services will have different requirements in terms of data rates and delay stability.
This means different management solutions have to be formulated for network optimization based on differentiated experiences. For example, optimizing for VR video services requires studying multiple cases and scenarios, and mapping different rates and delays to key network performance indicators (KPIs) such as SINR and RSRP.
Keep on optimizing
So the good news is that the challenges of 5G optimization aren’t insurmountable. The solutions aren’t just theoretical, either – Huawei says it’s been working closely with operators to help them ensure that their 5G rollouts are properly planned and optimized for maximum efficiency and network quality from the start. Huawei also says that as of mid-May 2019, its 5G ‘expert and tool platform’ optimization solution has been implemented by over 20 operators who are reporting significantly improved network capacity and usage rates.
However, while it’s good news that the above four challenges can be overcome, the not-so-good news is that network optimization challenges don’t stop there. For example, while 3G and 4G network optimization typically focuses on KPIs or KQIs (key quality indicators), the 5G era requires operators to optimize their networks to support service level agreements (SLAs).
The truth is that we’re going to see steady and rapid growth of many different types of 5G use cases and applications, many of which we haven’t even thought of yet. The more 5G enters the mainstream, the more new kinds of services will be introduced, which will make network optimization – and providing an optimal user experience for each new service – increasingly difficult. That means NSA 5G networks have to be optimized as they’re rolled out, not as an afterthought once the hardware and software are in place. By then, it could be too late.
This article is sponsored content from Huawei