Johannesburg, 01 Mar 2013
There is a growing gap between the amount of spectrum and network capacity available today, and the amount required to carry rising levels of mobile data worldwide. Mobile data traffic, dominated by video growth, will increase at least 20 times in volume between 2012 and 2017, creating huge traffic and signalling burdens for the wireless networks.
Traditional methods of increasing network capacity - upgrading the air interface and adding new base stations and spectrum - will be wholly inadequate to meet these challenges. In many cases, those remedies will deliver less than a third of the required additional capacity. While the introduction of LTE and the allocation of new spectrum licences are important, most of the capacity increase will come from a radical rethink of the network structure.
In particular, this will involve the use of millions of small cells to increase capacity and coverage to relieve the strain on the macro network. In future, these will evolve into heterogeneous networks, which will combine different layers of cells, supporting a variety of air interfaces and spectrum bands.
According to a recent survey of mobile carriers worldwide, conducted by Rethink Technology Research for Amdocs, over 80% of operators globally believe that small cells will be the first or second most important factor in meeting their capacity objectives between 2012 and 2017. However, despite their many advantages, these new approaches will also introduce a wide range of challenges in planning, managing and optimising the network. Despite rising levels of automation, carriers know that, to get the full business and performance benefits, they will need to invest in innovative tools and intelligence.
The rise of the small cell
Small cells started as indoor, private devices but are now starting to be introduced into the carriers' public networks, indoors or outdoors. The first to be deployed are typically single-mode - 3G or WiFi to start with. From 2013, there will be rising availability of, and trust in, multimode small cells, which can reduce cost and allow for integration and migration between 3G and 4G without swapping out the cells. Increasingly, WiFi will also be integrated in many units as a standard feature.
In its simplest form, a multimode cell just combines multiple radios and offers standard interfaces to the carrier network. But to use multimode products efficiently, new management capabilities will be introduced including enhanced scheduling, interference and power control, SON and intelligent mobility management, all taking the different available air interfaces into account for optimal performance. Further out, cognitive radio technologies may add to the cell's flexibility and responsiveness, but for now, most of these smarts are achieved in software.
Many carriers will opt for a heterogenous network, in which a macro layer provides wide umbrella coverage, and a separate layer of small cells, often in a different band, offers capacity hotzones. The traditional microcell is somewhat squeezed out of this picture over time. By 2017, no carrier expects to have a data network with microcells making up 50% or more of the total sites. By contrast, 15% will still have networks which are at least 50% composed of macrocells, but that figure will have been overtaken by systems with at least 50% small cells. The latter will be the case at one in five cellcos. There are still, however, many operators which are cautious about widespread small cell roll-outs, and by 2017 the most common topology will be a network with between 30% and 50% small cells, 20% to 30% microcells, and 20% to 50% macrocells.
Another important cell site issue is the integration of Wi-Fi into the base station. As well as moving rapidly towards networks which have a large component of small 3G/4G cells, operators are also deploying carrier WiFi, either as separate access points or integrated into cellular base stations, often as part of a HetNet. By 2017, only 12% expect to have no Wi-Fi integrated into their cellular networks, while 29% will have it incorporated in up to one-fifth of base stations, mainly small cells. About 40% think they will have WiFi integrated between 20% and 50% of cells by 2017.
Maximising the benefits of the new networks
With almost two-thirds of carriers expecting to see at least a tenfold increase in cell site numbers by 2017, they will face unprecedented challenges in terms of planning, performance measurement and management. They must ensure they do not just deliver additional capacity in an untargeted fashion, but in a way that delivers optimal cost-performance and supports key business objectives. The capabilities to plan and manage small cell networks will be enhanced as the standards evolve, but the standards will provide only a subset of the tools operators will require.
Traditionally, carriers would add capacity and speed to their networks by upgrading the air interface, and adding new spectrum and/or macrocell sites. Network planning to ensure greatest coverage and efficiency was complex but one-off, since RANs were not dynamic. Back end tools to analyse and prioritise different types of traffic or subscriber were in their infancy.
But in the mobile broadband era, a far wider range of techniques is necessary. While hardware and spectrum updates remain important, there will be a far greater emphasis on intelligence throughout the network to maximise capacity, efficiency and quality of service. This intelligence will be necessary to manage networks which are increasingly complex in terms of:
* Range of air interfaces and spectrum bands supported within one system
* Number and variety of cell sites
* Number and variety of backhaul connections
* Wide range of types of data traffic being handled
* Constantly varying levels of activity according to application, location and time of day
* Varying levels of quality of service according to application or subscription type
As traditional voice and messaging revenue streams decline, operators need a new profit model for data. A flexible, multi-technology network which can be managed in real time will be necessary to boost capacity while keeping costs down, and to support a wide range of new charging plans and applications, in order to retain customers, and to introduce new revenue streams like machine-to-machine.
All this will require a rethink of how the network is planned and managed, and the need to harness a whole range of tools and services.
Key planning and performance challenges
Operators have identified a range of planning challenges which will affect the performance and business returns of the new network. While some of these will be addressed to some extent by extensions to the standards, especially in LTE-Advanced, standards-based methods will address only part of the issue and additional tools will be required to gain optimal benefit from the small cell roll-outs.
The challenges can be broadly grouped into the following categories:
* Finding sites in the right position and planning their location relative to each other and other networks
* Backhaul
* Reporting effectively to support intelligent data delivery
* Integration with the macro and core networks without interference
As the graph shows, in the carrier survey, the five factors which were most commonly cited as the most crucial planning challenge all relate to cell sites, except one (integration with the core network). Securing sites in the best locations was rated as the most urgent challenge by 24% of respondents, followed by core integration (21%), the cost of acquiring small cell sites (19.5%) and the logistics of identifying and acquiring the sites (17%). Close behind came backhaul for small cell sites, cited by 12% as their most pressing challenge.
The single most decisive issue in ensuring that a small cell network delivers optimal benefits is the location of the cell sites. A recent study by Orange indicated that, in that operator's tests, the optimal distance from the macrocell is about 200 metres, with a small cell radius of about 30 metres. That can support 100% capacity gain for the macrocell, as well as improved quality of experience, while distances of under 50 meters add little capacity but plenty of interference. Other issues include the small cell's proximity to its neighbours and to cells run by other operators; and mapping the cell's position as accurately as possible to peak areas of data usage. The situation may be further complicated when several operators have cells in the same neighbourhood, with resulting interference risks.
All these considerations will create major planning challenges and several interrelated developments need to happen to instil full confidence about massive small cell roll-outs, according to the survey. The most commonly cited are:
* Planning and dimensioning tools specifically geared to small cells
* Mature self-organising and self-optimising network (SON) tools
* Standards for SON
* Affordable small cell backhaul options
* Legal framework for leasing small cell sites
As a result of all these concerns, 45% of operators expect to invest in new tools and services which are specifically optimised for small cell networks over the next four years. The new topology of the RANs will be a catalyst for new techniques, and for a shift in how networks are planned and measured, with a focus on methods that deliver high degrees of local accuracy, such as RNC data feeds.
Cost efficiencies
Of course, another key factor is that operators need to increase their RAN and backhaul capacity massively - by more than 50 times according to almost one-quarter of cellcos - but with capex and opex budgets that are under intense pressure. The typical operator will have only a 5%-10% increase in its backhaul capex budget, and up to 20% in the RAN, while opex constraints will be even tighter. These factors will make it essential to the business case that the new networks can be managed flexibly and efficiently, to deliver the greatest possible capacity where it is needed, and with a high level of automation - hence the intense interest in tools to reduce the manual overhead of running networks, or to allocate capacity more efficiently.
Indeed, a substantial number of operators are looking for an actual reduction in opex bills by using more modern technologies. 18% believe this is achievable in the RAN and 24% in backhaul. While 30% of carriers aim to keep their site opex stable and 18% to reduce it, while only 6% expect to spend more than 20% a year extra on RAN opex.
Conclusion
Carriers' budgets are under intense pressure and few will have the same kind of capex hikes which they enjoyed to deploy previous network upgrades. And opex targets are even more rigorous, with many carriers investing in new topologies precisely to save on running costs. Lower unit costs for cell sites and backhaul will only address part of the challenge, especially with the huge numbers envisaged. Instead, operators will have to rely on sophisticated and flexible planning approaches to maximise efficiency and automation, and to map capacity to returns.
Share