Radio planning tools are used throughout the wireless industry to determine the most effective location, configuration and parameters for base stations. But networks are becoming more complex and users are demanding better performance in the buildings where they work, live and visit. This means that the planning tools also need to become more sophisticated and work alongside automated SON (Self Organising Networks) technology built into the small cells and supporting servers.
Indoor data growth
Looking more closely at forecast data traffic growth, you'll see that an increasing percentage of data is consumed indoors. We tend to watch videos and surf the web on our tablets and smart phones sitting down, rather than on the move. Most of us also spend more of our time indoors.
Operators have traditionally designed their networks using outdoor/external cellsites to penetrate into buildings, believing this to be more cost effective than installing separate/independent indoor systems.
But as we move to the small cell architecture, the lower costs involved in deploying indoor small cells contrast strongly with outdoor. The units can be physically less robust because they operate in supervised environments and can be more easily installed; power is easily connected and they don't need wireless backhaul.
Separate indoor and outdoor worlds
Some operators deal with indoor, especially enterprise, and outdoor installations quite independently. Indoor system design and installation are often subcontracted to third parties who use their own tools, methods and installation teams.
Radio planning for inside buildings is often made using a separate RF planning tool specifically for the purpose. A sophisticated in-building RF planning tool normally takes a building as an input and designs network systems in the building. Typical features include building modelling from CAD files in dxf and dwg formats; network distribution system design; indoor radio propagation prediction with sub-meter resolution; and project management. Outdoor radio network planning tools take a map of an outdoor area as an input and design a network that provides service coverage. It normally includes a GIS module and an outdoor radio propagation prediction engine, typically with 20 to 100 meter resolution, which is not detailed enough to evaluate in-building network performance.
In an ideal world, separate dedicated spectrum will be allocated for small cell and indoor use, avoiding much of the issue of interference. No matter how much spectrum is available, operators will want to maximise its use and we can expect to share frequencies between indoor and outdoor systems in the long term.
As the number of indoor small cell installations increase, operators will want to evaluate the impact between indoor and outdoor networks and optimise them together, maximising the potential network performance from both.
There are few RF planning tools that address and co-ordinate between both indoor and outdoor environments, allowing these to be more closely aligned. One such product is RANPLAN's iBuildNet® tool which primarily focuses on in-building network design and optimisation, but also covers some high capacity indoor-outdoor scenarios such as campuses and CBDs (Central Business Districts) that comprise dozens of buildings, stadiums, airports and stations with high resolution - typically 20cm x 20cm x 20cm; fast and accurate coverage prediction; interference analysis and control; user throughput and network capacity analysis; and optimisation for 2G, 3G and LTE.
RF planning for an indoor installation looks at the signal distribution throughout the building on each floor. This is especially important for the high floors where there is often more interference from outdoor macrocells. The design of signal strength near the windows in particular reduces the chance of ping-pong handover to macrocells – where the handset toggles repeatedly between indoor and outdoor cellsites.
Lower floors in buildings have to control the signal leakage from the building to outdoors, otherwise there is a higher chance of outages and handover failures can arise. A useful target distance would be that at 10 metres from the building, 95% of the area would have a signal level of -95dBm.
Another factor to recognise is that 2G/3G/LTE and Wi-Fi, at both 2.4GHz and 5GHz, often co-exist within the same building and sometimes even at the same antenna. It's therefore important to be able to balance the uplink and downlink levels for all of these signals. Further complexity arises where MIMO and polarised antennas are used to enhance network capacity.
An operator might use an in-building RF planning tool primarily to evaluate the design from third parties and analyse feedback received about the quality of the network; from end user complaints, for example. The planning tool can be used to assess how and where problems arise, such as handover failures due to signal leakage outside from in-building systems. Where antennas are hidden inside ceilings or other less accessible places, it allows the design to be validated before installation and avoid costly modifications afterwards. An operator can also use an in-building RF planning tool to manage their indoor networks electronically in a central repository.
Those system integrators and equipment vendors that deliver the best quality designs will win more repeat business. Proper planning tools then make the system design and trunking/installation easier, and reduce the chance of remedial corrections later. The ability to co-ordinate between indoor and outdoor environments is key.
Comparing outdoor vs indoor small cells?
Some operators will prefer to use small cells on lampposts to penetrate indoors. First, we need to identify which particular type of buildings this works for – tall 50-floor buildings are out of the question; seven storeys are often considered as a reasonable maximum. Outdoor signals may not penetrate deep inside buildings, limiting both coverage and capacity. We also need to assess how much capacity can be delivered using an outside approach. Then we need to compare the performance with a dedicated in-building solution using several different scenarios.
The RF power level used by metrocells could be anything up to 5W – higher than that would be self-defeating, because it would reduce the overall network capacity. Metrocells larger than 1W are often much bulkier and unsuitable for lamppost mounting due to their size and environmental impact. This compares to indoor small cells that are often 250mW, with 1 Watt as a typical maximum.
How LTE/LTE-Advanced affects RF planning
There are a number of planning issues involved as we progress from 2G/3G/HSPA to LTE/LTE-Advanced.
Firstly, deploying LTE at 2.6GHz has a radically different coverage footprint from 900MHz or 1800MHz used for 2G/3G. The indoor penetration is much worse and handover time for LTE is shorter, reducing the room for error.
Secondly, MIMO (Multiple Input/Multiple Output) has to be modelled, which adds network complexity, particularly indoors.
Then, when we move to full HetNet, there is a much more complicated interference scenario to be considered. Traditional interference co-ordination won't work, which is why eICIC (enhanced Inter-Cell Interference Co-ordination) was developed.
Finally, when many small cells are deployed within the same coverage area, mobility management will be much more complicated. There will be greater loss of handover if the system is not designed well, resulting in outages, dropped calls or poor quality of service for users.
Practical constraints of LTE indoors
Studies of the use of higher order MIMO, particularly for in-building scenarios, evaluate which are the best MIMO configurations for maximum performance. It's a very complicated issue. For some in-building situations, there appears to be little benefit from using more than 2x2 MIMO. From a practical point of view, it could be expensive if using a passive DAS distribution system indoors and would mean a lot of extra cables.
Tightly co-ordinated HetNets (Heterogeneous Networks) using CoMP (Co-ordinated Multi-Point) can bring significant performance improvements at the edge of the cell coverage. This feature will be expensive to install, requiring closely aligned and co-ordinated remote radio head equipment rather than small cells. This extra cost may be justified in a few specific use cases where it wouldn't be easy to install small cells.
For example, imagine a very large public square with antenna surrounding the area but most of the traffic at the centre. These features should work very well in such a scenario. Football or soccer stadiums are also appropriate for this approach.
- The greater use of in-building small cells will create a need for better RF planning tools to ensure these are installed in the best locations.
- This needs a co-ordinated view across both indoor and outdoor environments, so that rather than using separate tools, a single combined RF planning tool will be needed.
- The adoption of SON (Self-Organising Networks) will help optimise the systems after installation, with RF planning tools used to help identify specific problem areas between indoor/outdoor co-ordination.
- The highest efficiency radio capacity solutions from LTE-Advanced will require specially designed and expensive HetNet systems. These may be appropriate for specific venues, but in most cases it will be cheaper and more viable to install a large numbers of small cells.
Our thanks to Prof Jie Zhang, Chair of Wireless Systems at Sheffield University for his help in writing this article.