North American operators are pushing forward with their LTE network deployments and, now that a number have laid their foundations, the focus must shift to ensuring the quality of those networks. But this is no easy task and, particularly from a network optimisation perspective, the road ahead is fraught with complexity.
When operators moved from 2G to 3G, there were a number of factors that the networks had in common, and there was clear continuity from a technical point of view. This is not the case as operators make their move to LTE, according to Frédéric Gonzales, managing director, sales and customer operations for the Americas region at Astellia. He explains the challenges facing operators as they embark upon LTE network deployment and how network optimisation solutions can help them address those challenges.
There are two broad differences between LTE networks and their 3G predecessors. LTE has a flat architectural structure, rather than hierarchical, and IP-based as the technology is designed to facilitate the efficient transfer of data packets. 3G networks, meanwhile, were originally designed for providing mobile voice, SMS and comparatively low-speed data. This architectural shift is a real game-changer in terms of network monitoring, according to Gonzales. He says that operators need to take time to understand the new structure and get to grips with using a network with a flat structure, and that it is important to deploy new technologies that are tailored specifically to an all-IP structure.
There is also an urgency associated with the move to LTE as there is less of a grace period for the technology to bed in. Consumers are much savvier about networks today than they were in the past, largely due to the widespread media hype around 4G networks and handsets. Such urgency was not apparent with 3G devices, and consumers are keen to get their hands on LTE-enabled devices to take advantage of faster data speeds.
“There is a high market pressure to deliver as fast as possible new devices to customers. It implies a shorter test period and so there is a possible impact on network performance and customer experience,” says Gonzales.
He adds that the same concern is applicable to the commercial release of new mobile apps. For instance, Apple’s video conferencing app Facetime is now usable over cellular networks, thanks to the firm’s upgrade to its new iOS6 platform. The app has already caused US operator AT&T to consider the impact it could have on the quality of its network. It is expected to be hugely popular among consumers and likely to place a heavy strain on the operator’s LTE network, which is why it has limited usage of it to a relatively small number of its premium subscribers.
And it is not just the proliferation of new smartphone apps that are expected to make operators’ lives difficult from a network optimisation perspective. Huge volumes of traffic in general will cause congestion issues, as smartphones come equipped with faster browsers, and content-rich websites and over-the-top services become more sophisticated to take advantage of new technologies on the market.
“It took almost eight years for the UMTS networks to hit major traffic growth,” says Gonzales. The first commercial deployment of a UMTS network was in 2001 when Japanese operator NTT DoCoMo rolled out its network, but Gonzales argues that managing the data boom only really became an issue for operators when the iPhone 3G arrived in 2008. “We are quite clearly not on the same timeframe for LTE. It means that optimisation concern will be a hot topic in the near future,” he says.
Add to the mix the complexity caused by the sheer diversity of spectrum available for LTE, and it is clear that ensuring a good experience on these networks for all users is a huge undertaking. There are 44 different bands allocated to LTE in use globally. In North America, the 700MHz and 800MHz bands, and the 1700MHz and 1900MHz bands are predominantly being used for LTE. However, in Europe, LTE networks are being used in the 800MHz, 900MHz, 1800MHz and 2.6GHz spectrum bands.
In comparison, there were just nine bands allocated to GSM across the world, of which only four are widely used. Added to that is the fact that there are two LTE technologies being deployed by operators: TDD-LTE and FDD-LTE networks. Therefore, optimising LTE networks for customers who roam internationally is a very important and challenging consideration.
It is therefore easy to see that operators that can provide a consistent, reliable and high quality experience stand to gain the edge over their rivals, and benefit from better brand reputation and loyalty if they deploy network optimisation solutions that are tailored to LTE networks.
But operators should not forget the one constant that runs through all generations of cellular technology; mobility. “For LTE, despite the fact that it is based on IP protocols and operators are able to optimise and monitor their networks by monitoring their IP activity, the key is still mobility for the customer and the handset in your network,” says Gonzales.
He classes the most important key performance indicators (KPIs) to monitor quality of experience for LTE networks into three main categories: Accessibility; ensuring that customers are able to access the network is an obvious metric to keep an eye on, so HTTP efficiency and bearer activation are key KPIs. A second metric is retainability, meaning that operators need to make sure they don’t interrupt customers’ experiences through dropped calls and data sessions. Activities such as tracking area updates and hand-over efficiency are useful in monitoring this. And finally, quality; making sure customers have the adequate throughput and latency that they expect from their service. Timer KPIs are also very important to this end, as they depict what subscribers hate the most, waiting.
In order to monitor these KPIs effectively, Astellia has developed the NOVA analytics solution. NOVA is a vendor independent solution that gives operators the waiting time from a customer point of view. The firm is presenting a demonstration of its NOVA Analytics solution for the first time on its booth during the LTE Summit in Dallas.
Gonzales says that the NOVA solution is aimed to ease such issues for its customers. He explains that the solution helps operators shift from the mindset of focusing on quality of service (QoS) and evolve towards QoE instead. In an all-IP world, the clarity of voice calls is no longer the focus, the efficient transfer of data is.
“4G networks are data-driven, so QoS strategies focus now mainly on application and user, instead of network element performance,” he explains. “The NOVA solution offers embedded deep packet inspection (DPI) capability to monitor application and services, but it also has an analytic solution with multi dimension filtering. This means that the added value that the solution offers is to gather three independent layers: user plane content, session signalling and mobility, to be able to optimise mobility of the handsets in an operator’s network when a customer moves from one node in the network to another and when they move to LTE to 3G because of lack of network coverage in a specific area.”
“We are offering a list of hundreds KPIs and KQIs already standardised in our NOVA solution to help the operator. The effectiveness of this has been validated by early deployments in the field.”
The firm’s offering also helps operators manage technical elements that many are not yet familiar with, such as OFDMA, which is the evolution of OFDM; the method of encoding data on multiple operators’ frequencies. Gonzales claims that OFDMA is mostly unknown to network engineers with experience in 2G and 3G and that it has so far only been used in Wimax networks.
“Most engineers don’t have knowledge of optimising this type of radio coding,” he says. “Thanks to our participation in early stage lab and field tests trials, we can offer operators expertise and even a complete managed services approach for them to optimise these kinds of complicated elements.”
LTE is not the only new technology that operators are deploying. Self Organised Networks (SONs) is another promising technology that is still in its nascent stages, and could make an operator’s life a lot easier in the long term. LTE SONs make use of network intelligence and management features in order to automate the configuration and optimisation of networks, lowering costs and improving network performance and flexibility.
This is taking up a lot of time and investment from operators that choose to deploy SONs and because it is not a mature technology, operators need to monitor their SONs carefully to minimise the impact on QoE.
“The implementation of SON technology is an important challenge from a QoE perspective, because we all know that introducing a new technology is always a challenge. The recommended approach is so to use vendor independent probe solution such as NOVA to monitor SON procedures’ related KPIs,” says Gonzales.
And a final obstacle when moving to LTE is the absence of the Iub interface, which has been critical in the monitoring and optimisation of networks in the past, he argues. Iub is the interface that connects nodes in the network to base stations, and there is no equivalent interface in LTE networks.
“Getting access to radio measurements is a huge challenge for engineers with LTE networks,” says Gonzales.
“With NOVA, this can be overcome. The solution uses X2 interfaces between nodes. A lot of information transits through this new interface, including some radio measurements. This interface is crucial to optimise mobility in the network and monitor SON procedures. We have found a solution to this crucial issue.”
Frédéric Gonzales, managing director, sales and customer operations for the Americas region at Astellia, talks about the firm’s approach to LTE optimisation.