Picocells Save You Money
Content
Picocells save you money
7 ways to avoid over-spending on your network
Chris Cox, Director of Marketing Programmes, ip.access May 2009
Executive summary
In the current economic climate there is increasing pressure on mobile operators to save costs and improve efficiency in all areas. In this white paper we outline seven different areas where picocells are already being used by operators around the world to save money on both the capital and operational costs of their networks. We’ll look at how picocells can help you… 1. Use less DAS in buildings 2. Save on roof access costs 3. Cut transmission costs 4. Slash the cost of temporary coverage 5. Offload the macro network - the low-cost way 6. Stop over-provisioning your 2G network - sweat your 2G assets 7. Cut the overheads on in-building deployments. When you compare picocells with the traditional approach of using a distributed antenna system (DAS) to provide coverage inside a building, the cost savings are clear. Picocells connect into the core network over an Ethernet cable, and as a result they eliminate the cabling and implementation costs of a DAS system, and can easily be moved if the customer changes location later. If a DAS system is already installed in the building, a picocell provides a very cost effective way to feed it. Unlike repeaters (a common alternative feed), picocells do not require roof mounting, which saves on installation and planning costs. Another alternative feed, the micro basestation, typically requires expensive E1/T1 transmission lines for backhaul. This cost can be saved within months by switching to an IP picocell. Other situations where picocells have saved significant costs include provision of fast and effective coverage for one-off events; reducing the costs associated with improving macro network performance; and freeing existing network assets for re-use. The simplicity and speed of picocell installations can also reduce the operational overhead necessitated by inbuilding deployments. Not all of the savings will apply in every situation - some of them depend on specific cost structures in a particular market, such as the price of a leased line into the building. So in each area we have broken out the individual elements that make up the savings, enabling you to see which are important in the context of your network and customers. However, taken together these savings can be very significant. It is not unusual for a picocell deployment to cost half as much and take half the time of a more traditional approach to the same problem.
-1-
1. Use less DAS in buildings
When there is a requirement to get coverage inside a building, most mobile operators will normally consider using a distributed antenna system (DAS). These have proven to be effective over the years. However DAS systems are very expensive to deploy, often costing $100k or more, and require a lot of time and effort for each building. As a result they are typically only used on large, high profile projects where it is essential that coverage is provided, and only a limited number are implemented each year. Picocells provide an inexpensive approach for smaller buildings, opening up new areas of the market that cannot be served cost-effectively with DAS. What’s more, they have also been shown to be around 50% cheaper to deploy than traditional DAS systems even in large buildings. Case study Installation of a DAS system is a very invasive process A major western European carrier needed to provide coverage to a that requires a lot of negotiation with the building owner, bank in the capital city. The planning and precise engineering in order to ensure that building was eight floors high and the system works as expected once it is installed. Since DAS systems are expensive, it is important that installation is done correctly first time. This requires a lot of time and effort ensuring that the system will work properly once it is installed. The mobile operator must obtain building plans and negotiate with the building owners. Very often there can be extensive discussions with the architect about what types of antennas are allowed and where they can be located. In addition, it is necessary to get access to the equipment cabinets and cabling ducts to run the cables necessary for the system.
the cost of a distributed antenna system was estimated to be approximately 100,000 euros. This building was covered using picocells at a cost of only 60,000 euros. The time to revenue was less than half the estimated six months it would have taken to deploy the distributed antenna system.
In many cases a site visit will be required with a test transmitter and TEMS phone to make measurements of the radio propagation within the building. This information is captured and fed back into the initial design to ensure the antennas are located in the correct place to achieve the desired coverage. The need for a site visit adds significantly to the total cost of installing a DAS system. Access to the ceiling spaces on each floor is essential to install a DAS system, which can be very disruptive to the building’s occupants. Hospitals are a particularly difficult environment, as strict infection control requirements mean that accessing wall and roof spaces is a very time consuming and expensive business. By direct comparison, picocells avoid this expensive and time consuming process - for example one Eastern European carrier found that the time from agreement to getting the site on air is a maximum of 1 week with the ip.access nanoGSM® picocell, whereas DAS takes two months or more. In the current economic environment, business customers are increasingly likely to move offices. Picocells avoid the sunk costs of a DAS system; they can easily be unscrewed from the wall and moved if the customer changes location after installation. Picocell Lower installation costs and ability to move with the customer Saving
-2-
1.1 Passive DAS
For a passive DAS system, cabling represents a significant part of the total cost. Both the direct costs of the coax cable, which can amount to many dollars per foot, and the labour cost for pulling the cable, make the DAS installation expensive. In order to ensure an adequate link budget to overcome the losses and noise in the system between the signal source and each antenna, a typical DAS installation will have multiple antennas per floor. In comparison, a typical picocell installation requires only one or possibly two picocells per floor because there is no path loss between the antenna and receiver in the picocell (they are both housed in the same unit). This greatly reduces the amount of cabling and installation time for the picocell system. The following table summarises the main cost differences between a passive DAS and picocell implementation: Cost Passive antennas Using passive DAS The cost of a typical passive antenna is relatively low (tens of dollars). However, to ensure adequate coverage multiple antennas are used per floor. The cost of fireproof cable can be many dollars per foot. And since there are many antennas per floor, cabling represents a large part of the cost of a coax system. Costs depend on local labour rates, but can represent a large component of the total cost, particularly as it often involves working in the roof spaces on each floor and drilling through walls. Whether the passive DAS system is fed by a repeater or micro basestation, the RF amplifiers require significant power to overcome the cable losses. Due to the signal loss in the coax cable, additional amplifiers are needed for large buildings. Using picocells Antennas are integrated into the picocell, and only a few picocells are required for coverage. Picocells use standard CAT5 Ethernet cables for both power and backhaul. In most cases the existing cables in the building can be used. If existing Ethernet cables are used there is negligible cost. If new cables are required they can use existing Ethernet cable ducts. Picocells have a very low power requirement (12W each), and do not require a special equipment room. Picocell Saving No antennas required.
Coax cable
Ethernet cable is much cheaper than coax.
Pulling cable
Very low labour costs.
Power
No payment to building owner for power in equipment room. RF amplifiers are not required.
RF amplifiers
Picocells generate their own signal and connect using Ethernet, so there is no need for RF amplifiers.
Table 1: picocells vs. passive DAS
-3-
1.2 Active DAS
In contrast to traditional passive DAS systems, many new active systems are emerging onto the market. These systems typically have a higher capital cost, but are cheaper to deploy as they use Ethernet cable or fibre to distribute the RF within the building, and so save the capital and labour cost of coax cable. Like picocells, an active DAS system can take advantage of the existing wiring within the building, and in many cases the cost of the active DAS will be similar to an equivalent picocell deployment. However active DAS needs an RF source either from a repeater or microBTS which a picocell does not require, so there is a considerable saving using picocells. Picocell Picocells do not need a separate RF source of either repeater or microBTS Saving In fact, picocells can be used to feed an active distributed antenna system in exactly the same way as a micro basestation. One recent example where this was done is the Mandarin Oriental Hotel in Boston, where the active distributed antenna system is fed from an ip.access nanoGSM picocell. If greater capacity is required, the picocell deployment can be increased from a single TRX up to 4 TRX without loss in quality or performance.
1.3 Further considerations with DAS systems
In situations where many different mobile operators choose to share the same infrastructure, the cost saving of a picocell for one operator can be outweighed by the extra costs of deployments for the others. In these cases it may be better to use a traditional distributed antenna system rather than a picocell based system. In addition, a correctly designed DAS system can be shared by a number of radio technologies such as GSM and WCDMA, whereas new picocells would need to be deployed for each type of radio. However each of these technologies places its own specific requirements on the DAS system. For example, to ensure a good C/I to deliver HSPA data over WCDMA typically requires about 50% more antennas than a standard voiced based GSM system. Catering to possible future requirements therefore pushes up all the costs of cable, antennas and amplifiers, making the initial DAS deployment even more costly.
2. Save on roof access costs
Distributed antenna systems distribute the RF signal around the building, but they require a source of RF. DAS systems are typically fed by repeaters, using a donor antenna mounted on the roof of the building. There can be quite a significant cost associated with roof access (rental of the space on the roof) as well as the capital cost of the RF boosters needed to bring the signals from the roof down to the antennas inside the building. In some cases roof access may not be possible at all. One ip.access customer reported that, in their country, while banks would contribute to the cost of an in-building system they would not permit any antennas to be installed on the roof.
-4-
Cost Planning permission
Using a repeater In some countries, visible rooftop installations may require planning permission before installation, which can take many months. In many countries landlords have realised they can gain additional revenue by renting roof space to mobile operators.
Using picocells Picocells do not require planning permission.
Picocell Saving Saves cost and time associated with planning permits. No need for a rooftop lease, which can save as much as $1000 per month. No need for antenna installation.
Rooftop lease
Picocells do not need access to the rooftop.
Rooftop installation
Repeater hardware
Macro capacity
Interference management
Installation of the receiving antenna on the building roof may require skilled engineers, and permits from the building manager. Band selective repeaters are costly and require configuration. (Lower cost repeater hardware may retransmit competitors’ signals, which is illegal in many countries, and cause interference issues.) Subscribers in the building use voice and data capacity from the macro network, which may result in congestion and lead to an earlier requirement for macro network upgrades. Depending on the quality of installation and equipment, repeaters may generate interference for BTS sites other than the planned donor, especially following network replanning. These can be time consuming and costly for the RAN team to resolve.
No rooftop installation is required.
Each picocell can either cover a part of the building directly or can feed an antenna system.
No need for donor antenna and amplifier.
Picocells create extra capacity, and so reduce the loading on the macro network.
Offloads the macro network.
Picocells are directly managed from the NOC and have limited output power, thereby minimising interference issues and taking the hassle out of managing them.
Fewer interference issues and more direct control of the network.
Table 2: cost savings vs. rooftop repeaters
Overcoming all the issues to install a repeater and DAS system can take a long time. One operator reported that before the introduction of picocells they used repeaters which typically took 5-6 months to install. This was a big problem for business customers who would sometimes await coverage for the first 6 months of a 2 year contract, leading to dissatisfaction and churn. Picocells solved this problem.
-5-
3. Cut transmission costs
With today’s drive for efficiency, there is an increased focus on improving operating costs. One of the largest components is transmission, which can account for up to 40% of total operating costs for a mobile operator. Picocells can help to reduce this. Operators frequently use micro basestations to serve hotspots and feed DAS systems. These basestations connect into the core network over a leased E1 or T1 line. The cost of the leased line depends on where the building is (distance from the local exchange), with the exact nature of these costs varying significantly from country to country. In some countries the cost of an E1 line can be several thousand euros per year. Since picocells use IP they can be connected using a DSL line at a typical costs of a few hundred euros per year. Picocell Difference between cost of E1 / T1 lines and DSL Saving Case study
One western European operator was able to justify the cost of installing new ip.access picocells solely on the basis of the cost savings on E1 backhaul. They found that replacing an existing two-TRX micro basestation with a picocell would pay back the capital and installation costs of the picocell within fourteen months through transmission cost savings alone.
4. Slash the cost of temporary coverage
There are many situations where demand for mobile services in a particular location increases dramatically due to occasional events. For example, a race track or a festival ground may require greatly increased network coverage and capacity a few times each year. Providing capacity for such peaks can be very expensive, but ignoring the problem - simply allowing the cell to saturate during the period of the event - can lead to customer dissatisfaction. Carriers have a number of choices for satisfying the demand. The traditional approach to providing temporary additional capacity is to use a so-called CoW (cellular on wheels); a truck with a complete a macro basestation and mast. These CoWs are very expensive to run, and each deployment may cost many tens of thousands of euros. Since picocells are small, portable (the size of a laptop) and cheap, they can be carried to the required site very easily and deployed using any form of available IP connectivity. This can include existing broadband infrastructure at the venue, but also point-to-point wireless connections. At the end of the event the picocells can be recovered very easily and reused for another event at another time. Case study
ip.access picocells were used in the press office at the 2009 Monaco Grand Prix to enable journalists to make and receive phone calls and to support their use of BlackBerries for email during the event.
In this way, picocells alleviate the need to run multiple CoWs, with their associated crews kept trained and ready to operate. This dramatically reduces the average cost to cover a one-off event. While picocells do not provide the output power and range of a typical CoW they do add significant additional capacity to the network, and they allow coverage to be targeted exactly where it is needed. Picocell Reduced need for expensive temporary coverage trucks Saving
-6-
5. Offload the macro network - the low-cost way
For most operators the 2G network is a significant asset that is heavily used, but there is a strategic desire to reduce further investment in it. As a result, some cell sites may become heavily loaded and experience operational difficulties such as high call failure rates. Existing BTSs can be upgraded, but in some cases the only solution is acquisition of a new BTS site. In many developed markets this is a very time consuming and expensive process which can take years to gain planning approval. In some countries half of all planning applications fail, and in others new sites are almost impossible to obtain in city centres, so different approaches to providing GSM coverage have to be adopted. Picocells give the operator a new way of solving this problem. By deploying picocells inside buildings within the coverage footprint of the macrocell, the operator can move traffic from a heavily congested site onto the picocell layer. This alleviates the overloading of the macrocell and provides a high quality of service to those customers where the picocell is deployed. Studies show that approximately 70% of mobile phone traffic is generated by users inside buildings so there is ample scope for offloading traffic from congested cells. At the same time this approach can be used to provide a very high quality of service to key customers – for example in enterprise premises.
Case Study T-Mobile US began using nanoGSM picocells from ip.access because of their ease of installation, low cost and use of IP connections for backhaul. Yvonne Manns, T-Mobile’s Senior Regional Radio Frequency Engineering Manager for In-building Solutions for the Northeast Region says, “When a customer needs additional inbuilding coverage, we can deliver it really quickly since the IP unit is small, inexpensive and utilizes a cheaper backhaul solution which helps to reduce OPEX costs.” The ip.access nanoGSM BTS removes the problems of cell distortion, interference, handover and management in city centre locations where macro BTS spacing can be as low as 200 metres. “They provide extra capacity indoors to ensure that our subscribers receive the radio quality they need to access our network and to maintain their call,” adds Manns. “Picocells can be deployed very cost effectively and efficiently, which helps to retain our Enterprise customers in need of additional coverage. They also help to improve our macro network’s spectral efficiency and enhance the macro sectors’ call quality. That’s why we use them.”
Some carriers take this approach further and use picocells to provide coverage outdoors in some urban areas. There are a number of different ways this has been done; for example the picocell can be mounted in an outdoor enclosure fixed to street furniture such as a lamp post, or alternatively the picocell can be deployed inside a building to provide coverage outdoors using an external antenna. In this way the operator is able to add capacity to the outdoor network in areas where this would otherwise be impossible. By shifting traffic onto an in-building network, the operator is able to improve the overall performance of a congested cell. Since picocells cost very little to deploy compared to a macro basestation upgrade or a new macro basestation, they provide a very quick and costeffective way of improving service to all network users. Picocell Multiple picocells can be deployed more quickly and cheaply than the lengthy site Saving acquisition and deployment of a new macro BTS
-7-
6. Stop over-provisioning – sweat your 2G assets
As users migrate from 2G to 3G the pattern of network usage changes. This can leave some areas relatively oversupplied with expensive micro basestations providing more capacity than is required, while in other areas the use of 2G is still growing. Since picocells are quick and simple to deploy, they can be used to replace more expensive micro basestations where the higher capacity is no longer needed. The picocell solution has lower capital and operating costs, and frees up the micro basestations to be used elsewhere in the network where the extra capacity is still needed. By sweating their assets in this way, operators can minimise their investment in GSM technology and make their existing assets work much harder. Picocell Postpone the purchase of new micro BTSs Saving As operators review their use of spectrum, they may wish to move users from one band to another. For example, many European operators are considering moving GSM users to the 1800 MHz band to clear 900 MHz spectrum for UMTS. Given the higher frequency of the 1800 MHz band this may result in some new coverage holes on the GSM network, which will increase customer complaints and could potentially cause churn. However, operators would generally prefer not to make significant investments in upgrading their macro GSM basestations to resolve this issue. Picocells can be used to plug the gaps in a more targeted and cost-effective way. Picocell Remove barriers to refarming of 900 MHz spectrum for UMTS Saving
7. Cut the overheads on in-building deployments
Overhead costs are hard to quantify, but they occur in each phase of an in-building system deployment, and limit the scope for more efficient operations. The design of a traditional in-building system requires skill and experience, consuming valuable resources within the radio planning team, and may require extensive negotiations with the building owners and architects. Once the design has been completed and cost estimates obtained, most major in-building systems undergo a business case review cycle involving sales and finance staff. This decision-making process again requires time from key people within the business. After a decision to go ahead has been made, the radio planning teams will typically be required to supervise contractors who are implementing the in-building system. They may also be involved in testing and certification to ensure that the implementation has met its design requirements. The team will be required to work with the contractor to resolve any issues. All of these hidden resource costs limit the number of in-building projects that the radio planning team can deliver each year. As picocells are straightforward to install and test, some mobile operators have chosen to work with contractors who offer a fixed price for specific types of deployment. This eliminates the overhead associated with design, decision-making and installation of an in-building system, because everything is included in the fixed price. It also reduces the need for internal discussions and detailed business case development, as the pre-sales and sales teams know
-8-
in advance what the cost will be, freeing resources to concentrate on selling and optimizing the network. Resolving customer complaints about the network is another area which can generate significant overhead. First the complaint has to be investigated and then the solution has to be worked out, which requires the attention of highly qualified people. All this takes time, during which the unhappy customer may be lost to a rival network. Since picocells are so quick to deploy, field engineers investigating a customer complaint can carry one with them when they go to visit the customer. If appropriate, they are then in a position to install the picocell (using the customer’s own broadband connection) to resolve the complaint on the spot. This reduces the number of visits required, bringing down the cost of complaint resolution and significantly improving customer satisfaction. Picocell Lower overheads and quicker complaint resolution from an efficient process for inSaving building deployments
Conclusion
We’ve looked at seven different ways to drive down costs in your network using picocells, so you can invest where the growth is: 1. Use less DAS in buildings 2. Save on roof access costs 3. Cut transmission costs 4. Slash the cost of temporary coverage 5. Offload the macro network - the low-cost way 6. Stop over-provisioning your 2G network - sweat your 2G assets 7. Cut the overheads on in-building deployments. But we haven’t even included one rather obvious cost - that of losing business customers because of bad coverage or limited network capacity. Coverage is a significant reason for churn. The lifetime value of each customer you save by plugging in a picocell pays for the kit many times over. We didn’t include this because you may not have any unhappy business customers (but then again, you may not have those customers for a reason). The bottom line is that picocells are a surgical instrument for removing costs from your network without sacrificing coverage, capacity or customer satisfaction. In fact, all three of these things improve markedly with picocells.
-9-
About ip.access
Based in Cambridge, UK, ip.access ltd. is a leading manufacturer of cost-effective picocell and femtocell infrastructure solutions for GSM, GPRS, EDGE and 3G. These solutions bring IP and cellular technologies together drive down costs and increase coverage and capacity of mobile networks. ip.access is the company behind the multi-award winning Oyster 3G™ femtocell, which dramatically improves the 3G experience at home. The ip.access nanoGSM® is the world’s most deployed picocell solution, carrying over a billion minutes of voice traffic every year for operators all over the world, including T-Mobile, TeliaSonera, and Telefónica O2. It provides GSM, GPRS and EDGE coverage and capacity for offices, shops and using satellite backhaul passenger aircraft, ships and in remote rural areas. With deployments in more than 40 live networks around the world and growing, ip.access is the partner of choice for operators competing in the new converged marketplace. ip.access is an active member of the Femto Forum, ETSI and the Network Vendors Interoperability Testing Forum.
[email protected]
Copyright © ip.access 2009. Oyster 3G™, nanoGSM® and nano3G™ are trademarks of ip.access ltd. All other trademarks are acknowledged. This document contains advance information, subject to change without notice. No responsibility is assumed by ip.access for the use of this information, nor for infringements of patents or other rights of third parties. This document is the property of ip.access and implies no license under patents, copyrights or trade secrets. No part of this publication may be copied, reproduced, stored in a retrieval system, or transmitted, in any form of any means, electronic, photographic, or otherwise, or used as the basis for manufacture or sale of any items without the prior written consent of ip.access.
- 10 -
ip.access ltd Building 2020 ■ Cambourne Business Park ■ Cambourne ■ Cambridge ■ CB23 6DW ■ UK T +44(0)1954 713700 ■ F +44(0)1954 713799 ■ [email protected] www.ipaccess.com
- 11 -
7 ways to avoid over-spending on your network
Chris Cox, Director of Marketing Programmes, ip.access May 2009
Executive summary
In the current economic climate there is increasing pressure on mobile operators to save costs and improve efficiency in all areas. In this white paper we outline seven different areas where picocells are already being used by operators around the world to save money on both the capital and operational costs of their networks. We’ll look at how picocells can help you… 1. Use less DAS in buildings 2. Save on roof access costs 3. Cut transmission costs 4. Slash the cost of temporary coverage 5. Offload the macro network - the low-cost way 6. Stop over-provisioning your 2G network - sweat your 2G assets 7. Cut the overheads on in-building deployments. When you compare picocells with the traditional approach of using a distributed antenna system (DAS) to provide coverage inside a building, the cost savings are clear. Picocells connect into the core network over an Ethernet cable, and as a result they eliminate the cabling and implementation costs of a DAS system, and can easily be moved if the customer changes location later. If a DAS system is already installed in the building, a picocell provides a very cost effective way to feed it. Unlike repeaters (a common alternative feed), picocells do not require roof mounting, which saves on installation and planning costs. Another alternative feed, the micro basestation, typically requires expensive E1/T1 transmission lines for backhaul. This cost can be saved within months by switching to an IP picocell. Other situations where picocells have saved significant costs include provision of fast and effective coverage for one-off events; reducing the costs associated with improving macro network performance; and freeing existing network assets for re-use. The simplicity and speed of picocell installations can also reduce the operational overhead necessitated by inbuilding deployments. Not all of the savings will apply in every situation - some of them depend on specific cost structures in a particular market, such as the price of a leased line into the building. So in each area we have broken out the individual elements that make up the savings, enabling you to see which are important in the context of your network and customers. However, taken together these savings can be very significant. It is not unusual for a picocell deployment to cost half as much and take half the time of a more traditional approach to the same problem.
-1-
1. Use less DAS in buildings
When there is a requirement to get coverage inside a building, most mobile operators will normally consider using a distributed antenna system (DAS). These have proven to be effective over the years. However DAS systems are very expensive to deploy, often costing $100k or more, and require a lot of time and effort for each building. As a result they are typically only used on large, high profile projects where it is essential that coverage is provided, and only a limited number are implemented each year. Picocells provide an inexpensive approach for smaller buildings, opening up new areas of the market that cannot be served cost-effectively with DAS. What’s more, they have also been shown to be around 50% cheaper to deploy than traditional DAS systems even in large buildings. Case study Installation of a DAS system is a very invasive process A major western European carrier needed to provide coverage to a that requires a lot of negotiation with the building owner, bank in the capital city. The planning and precise engineering in order to ensure that building was eight floors high and the system works as expected once it is installed. Since DAS systems are expensive, it is important that installation is done correctly first time. This requires a lot of time and effort ensuring that the system will work properly once it is installed. The mobile operator must obtain building plans and negotiate with the building owners. Very often there can be extensive discussions with the architect about what types of antennas are allowed and where they can be located. In addition, it is necessary to get access to the equipment cabinets and cabling ducts to run the cables necessary for the system.
the cost of a distributed antenna system was estimated to be approximately 100,000 euros. This building was covered using picocells at a cost of only 60,000 euros. The time to revenue was less than half the estimated six months it would have taken to deploy the distributed antenna system.
In many cases a site visit will be required with a test transmitter and TEMS phone to make measurements of the radio propagation within the building. This information is captured and fed back into the initial design to ensure the antennas are located in the correct place to achieve the desired coverage. The need for a site visit adds significantly to the total cost of installing a DAS system. Access to the ceiling spaces on each floor is essential to install a DAS system, which can be very disruptive to the building’s occupants. Hospitals are a particularly difficult environment, as strict infection control requirements mean that accessing wall and roof spaces is a very time consuming and expensive business. By direct comparison, picocells avoid this expensive and time consuming process - for example one Eastern European carrier found that the time from agreement to getting the site on air is a maximum of 1 week with the ip.access nanoGSM® picocell, whereas DAS takes two months or more. In the current economic environment, business customers are increasingly likely to move offices. Picocells avoid the sunk costs of a DAS system; they can easily be unscrewed from the wall and moved if the customer changes location after installation. Picocell Lower installation costs and ability to move with the customer Saving
-2-
1.1 Passive DAS
For a passive DAS system, cabling represents a significant part of the total cost. Both the direct costs of the coax cable, which can amount to many dollars per foot, and the labour cost for pulling the cable, make the DAS installation expensive. In order to ensure an adequate link budget to overcome the losses and noise in the system between the signal source and each antenna, a typical DAS installation will have multiple antennas per floor. In comparison, a typical picocell installation requires only one or possibly two picocells per floor because there is no path loss between the antenna and receiver in the picocell (they are both housed in the same unit). This greatly reduces the amount of cabling and installation time for the picocell system. The following table summarises the main cost differences between a passive DAS and picocell implementation: Cost Passive antennas Using passive DAS The cost of a typical passive antenna is relatively low (tens of dollars). However, to ensure adequate coverage multiple antennas are used per floor. The cost of fireproof cable can be many dollars per foot. And since there are many antennas per floor, cabling represents a large part of the cost of a coax system. Costs depend on local labour rates, but can represent a large component of the total cost, particularly as it often involves working in the roof spaces on each floor and drilling through walls. Whether the passive DAS system is fed by a repeater or micro basestation, the RF amplifiers require significant power to overcome the cable losses. Due to the signal loss in the coax cable, additional amplifiers are needed for large buildings. Using picocells Antennas are integrated into the picocell, and only a few picocells are required for coverage. Picocells use standard CAT5 Ethernet cables for both power and backhaul. In most cases the existing cables in the building can be used. If existing Ethernet cables are used there is negligible cost. If new cables are required they can use existing Ethernet cable ducts. Picocells have a very low power requirement (12W each), and do not require a special equipment room. Picocell Saving No antennas required.
Coax cable
Ethernet cable is much cheaper than coax.
Pulling cable
Very low labour costs.
Power
No payment to building owner for power in equipment room. RF amplifiers are not required.
RF amplifiers
Picocells generate their own signal and connect using Ethernet, so there is no need for RF amplifiers.
Table 1: picocells vs. passive DAS
-3-
1.2 Active DAS
In contrast to traditional passive DAS systems, many new active systems are emerging onto the market. These systems typically have a higher capital cost, but are cheaper to deploy as they use Ethernet cable or fibre to distribute the RF within the building, and so save the capital and labour cost of coax cable. Like picocells, an active DAS system can take advantage of the existing wiring within the building, and in many cases the cost of the active DAS will be similar to an equivalent picocell deployment. However active DAS needs an RF source either from a repeater or microBTS which a picocell does not require, so there is a considerable saving using picocells. Picocell Picocells do not need a separate RF source of either repeater or microBTS Saving In fact, picocells can be used to feed an active distributed antenna system in exactly the same way as a micro basestation. One recent example where this was done is the Mandarin Oriental Hotel in Boston, where the active distributed antenna system is fed from an ip.access nanoGSM picocell. If greater capacity is required, the picocell deployment can be increased from a single TRX up to 4 TRX without loss in quality or performance.
1.3 Further considerations with DAS systems
In situations where many different mobile operators choose to share the same infrastructure, the cost saving of a picocell for one operator can be outweighed by the extra costs of deployments for the others. In these cases it may be better to use a traditional distributed antenna system rather than a picocell based system. In addition, a correctly designed DAS system can be shared by a number of radio technologies such as GSM and WCDMA, whereas new picocells would need to be deployed for each type of radio. However each of these technologies places its own specific requirements on the DAS system. For example, to ensure a good C/I to deliver HSPA data over WCDMA typically requires about 50% more antennas than a standard voiced based GSM system. Catering to possible future requirements therefore pushes up all the costs of cable, antennas and amplifiers, making the initial DAS deployment even more costly.
2. Save on roof access costs
Distributed antenna systems distribute the RF signal around the building, but they require a source of RF. DAS systems are typically fed by repeaters, using a donor antenna mounted on the roof of the building. There can be quite a significant cost associated with roof access (rental of the space on the roof) as well as the capital cost of the RF boosters needed to bring the signals from the roof down to the antennas inside the building. In some cases roof access may not be possible at all. One ip.access customer reported that, in their country, while banks would contribute to the cost of an in-building system they would not permit any antennas to be installed on the roof.
-4-
Cost Planning permission
Using a repeater In some countries, visible rooftop installations may require planning permission before installation, which can take many months. In many countries landlords have realised they can gain additional revenue by renting roof space to mobile operators.
Using picocells Picocells do not require planning permission.
Picocell Saving Saves cost and time associated with planning permits. No need for a rooftop lease, which can save as much as $1000 per month. No need for antenna installation.
Rooftop lease
Picocells do not need access to the rooftop.
Rooftop installation
Repeater hardware
Macro capacity
Interference management
Installation of the receiving antenna on the building roof may require skilled engineers, and permits from the building manager. Band selective repeaters are costly and require configuration. (Lower cost repeater hardware may retransmit competitors’ signals, which is illegal in many countries, and cause interference issues.) Subscribers in the building use voice and data capacity from the macro network, which may result in congestion and lead to an earlier requirement for macro network upgrades. Depending on the quality of installation and equipment, repeaters may generate interference for BTS sites other than the planned donor, especially following network replanning. These can be time consuming and costly for the RAN team to resolve.
No rooftop installation is required.
Each picocell can either cover a part of the building directly or can feed an antenna system.
No need for donor antenna and amplifier.
Picocells create extra capacity, and so reduce the loading on the macro network.
Offloads the macro network.
Picocells are directly managed from the NOC and have limited output power, thereby minimising interference issues and taking the hassle out of managing them.
Fewer interference issues and more direct control of the network.
Table 2: cost savings vs. rooftop repeaters
Overcoming all the issues to install a repeater and DAS system can take a long time. One operator reported that before the introduction of picocells they used repeaters which typically took 5-6 months to install. This was a big problem for business customers who would sometimes await coverage for the first 6 months of a 2 year contract, leading to dissatisfaction and churn. Picocells solved this problem.
-5-
3. Cut transmission costs
With today’s drive for efficiency, there is an increased focus on improving operating costs. One of the largest components is transmission, which can account for up to 40% of total operating costs for a mobile operator. Picocells can help to reduce this. Operators frequently use micro basestations to serve hotspots and feed DAS systems. These basestations connect into the core network over a leased E1 or T1 line. The cost of the leased line depends on where the building is (distance from the local exchange), with the exact nature of these costs varying significantly from country to country. In some countries the cost of an E1 line can be several thousand euros per year. Since picocells use IP they can be connected using a DSL line at a typical costs of a few hundred euros per year. Picocell Difference between cost of E1 / T1 lines and DSL Saving Case study
One western European operator was able to justify the cost of installing new ip.access picocells solely on the basis of the cost savings on E1 backhaul. They found that replacing an existing two-TRX micro basestation with a picocell would pay back the capital and installation costs of the picocell within fourteen months through transmission cost savings alone.
4. Slash the cost of temporary coverage
There are many situations where demand for mobile services in a particular location increases dramatically due to occasional events. For example, a race track or a festival ground may require greatly increased network coverage and capacity a few times each year. Providing capacity for such peaks can be very expensive, but ignoring the problem - simply allowing the cell to saturate during the period of the event - can lead to customer dissatisfaction. Carriers have a number of choices for satisfying the demand. The traditional approach to providing temporary additional capacity is to use a so-called CoW (cellular on wheels); a truck with a complete a macro basestation and mast. These CoWs are very expensive to run, and each deployment may cost many tens of thousands of euros. Since picocells are small, portable (the size of a laptop) and cheap, they can be carried to the required site very easily and deployed using any form of available IP connectivity. This can include existing broadband infrastructure at the venue, but also point-to-point wireless connections. At the end of the event the picocells can be recovered very easily and reused for another event at another time. Case study
ip.access picocells were used in the press office at the 2009 Monaco Grand Prix to enable journalists to make and receive phone calls and to support their use of BlackBerries for email during the event.
In this way, picocells alleviate the need to run multiple CoWs, with their associated crews kept trained and ready to operate. This dramatically reduces the average cost to cover a one-off event. While picocells do not provide the output power and range of a typical CoW they do add significant additional capacity to the network, and they allow coverage to be targeted exactly where it is needed. Picocell Reduced need for expensive temporary coverage trucks Saving
-6-
5. Offload the macro network - the low-cost way
For most operators the 2G network is a significant asset that is heavily used, but there is a strategic desire to reduce further investment in it. As a result, some cell sites may become heavily loaded and experience operational difficulties such as high call failure rates. Existing BTSs can be upgraded, but in some cases the only solution is acquisition of a new BTS site. In many developed markets this is a very time consuming and expensive process which can take years to gain planning approval. In some countries half of all planning applications fail, and in others new sites are almost impossible to obtain in city centres, so different approaches to providing GSM coverage have to be adopted. Picocells give the operator a new way of solving this problem. By deploying picocells inside buildings within the coverage footprint of the macrocell, the operator can move traffic from a heavily congested site onto the picocell layer. This alleviates the overloading of the macrocell and provides a high quality of service to those customers where the picocell is deployed. Studies show that approximately 70% of mobile phone traffic is generated by users inside buildings so there is ample scope for offloading traffic from congested cells. At the same time this approach can be used to provide a very high quality of service to key customers – for example in enterprise premises.
Case Study T-Mobile US began using nanoGSM picocells from ip.access because of their ease of installation, low cost and use of IP connections for backhaul. Yvonne Manns, T-Mobile’s Senior Regional Radio Frequency Engineering Manager for In-building Solutions for the Northeast Region says, “When a customer needs additional inbuilding coverage, we can deliver it really quickly since the IP unit is small, inexpensive and utilizes a cheaper backhaul solution which helps to reduce OPEX costs.” The ip.access nanoGSM BTS removes the problems of cell distortion, interference, handover and management in city centre locations where macro BTS spacing can be as low as 200 metres. “They provide extra capacity indoors to ensure that our subscribers receive the radio quality they need to access our network and to maintain their call,” adds Manns. “Picocells can be deployed very cost effectively and efficiently, which helps to retain our Enterprise customers in need of additional coverage. They also help to improve our macro network’s spectral efficiency and enhance the macro sectors’ call quality. That’s why we use them.”
Some carriers take this approach further and use picocells to provide coverage outdoors in some urban areas. There are a number of different ways this has been done; for example the picocell can be mounted in an outdoor enclosure fixed to street furniture such as a lamp post, or alternatively the picocell can be deployed inside a building to provide coverage outdoors using an external antenna. In this way the operator is able to add capacity to the outdoor network in areas where this would otherwise be impossible. By shifting traffic onto an in-building network, the operator is able to improve the overall performance of a congested cell. Since picocells cost very little to deploy compared to a macro basestation upgrade or a new macro basestation, they provide a very quick and costeffective way of improving service to all network users. Picocell Multiple picocells can be deployed more quickly and cheaply than the lengthy site Saving acquisition and deployment of a new macro BTS
-7-
6. Stop over-provisioning – sweat your 2G assets
As users migrate from 2G to 3G the pattern of network usage changes. This can leave some areas relatively oversupplied with expensive micro basestations providing more capacity than is required, while in other areas the use of 2G is still growing. Since picocells are quick and simple to deploy, they can be used to replace more expensive micro basestations where the higher capacity is no longer needed. The picocell solution has lower capital and operating costs, and frees up the micro basestations to be used elsewhere in the network where the extra capacity is still needed. By sweating their assets in this way, operators can minimise their investment in GSM technology and make their existing assets work much harder. Picocell Postpone the purchase of new micro BTSs Saving As operators review their use of spectrum, they may wish to move users from one band to another. For example, many European operators are considering moving GSM users to the 1800 MHz band to clear 900 MHz spectrum for UMTS. Given the higher frequency of the 1800 MHz band this may result in some new coverage holes on the GSM network, which will increase customer complaints and could potentially cause churn. However, operators would generally prefer not to make significant investments in upgrading their macro GSM basestations to resolve this issue. Picocells can be used to plug the gaps in a more targeted and cost-effective way. Picocell Remove barriers to refarming of 900 MHz spectrum for UMTS Saving
7. Cut the overheads on in-building deployments
Overhead costs are hard to quantify, but they occur in each phase of an in-building system deployment, and limit the scope for more efficient operations. The design of a traditional in-building system requires skill and experience, consuming valuable resources within the radio planning team, and may require extensive negotiations with the building owners and architects. Once the design has been completed and cost estimates obtained, most major in-building systems undergo a business case review cycle involving sales and finance staff. This decision-making process again requires time from key people within the business. After a decision to go ahead has been made, the radio planning teams will typically be required to supervise contractors who are implementing the in-building system. They may also be involved in testing and certification to ensure that the implementation has met its design requirements. The team will be required to work with the contractor to resolve any issues. All of these hidden resource costs limit the number of in-building projects that the radio planning team can deliver each year. As picocells are straightforward to install and test, some mobile operators have chosen to work with contractors who offer a fixed price for specific types of deployment. This eliminates the overhead associated with design, decision-making and installation of an in-building system, because everything is included in the fixed price. It also reduces the need for internal discussions and detailed business case development, as the pre-sales and sales teams know
-8-
in advance what the cost will be, freeing resources to concentrate on selling and optimizing the network. Resolving customer complaints about the network is another area which can generate significant overhead. First the complaint has to be investigated and then the solution has to be worked out, which requires the attention of highly qualified people. All this takes time, during which the unhappy customer may be lost to a rival network. Since picocells are so quick to deploy, field engineers investigating a customer complaint can carry one with them when they go to visit the customer. If appropriate, they are then in a position to install the picocell (using the customer’s own broadband connection) to resolve the complaint on the spot. This reduces the number of visits required, bringing down the cost of complaint resolution and significantly improving customer satisfaction. Picocell Lower overheads and quicker complaint resolution from an efficient process for inSaving building deployments
Conclusion
We’ve looked at seven different ways to drive down costs in your network using picocells, so you can invest where the growth is: 1. Use less DAS in buildings 2. Save on roof access costs 3. Cut transmission costs 4. Slash the cost of temporary coverage 5. Offload the macro network - the low-cost way 6. Stop over-provisioning your 2G network - sweat your 2G assets 7. Cut the overheads on in-building deployments. But we haven’t even included one rather obvious cost - that of losing business customers because of bad coverage or limited network capacity. Coverage is a significant reason for churn. The lifetime value of each customer you save by plugging in a picocell pays for the kit many times over. We didn’t include this because you may not have any unhappy business customers (but then again, you may not have those customers for a reason). The bottom line is that picocells are a surgical instrument for removing costs from your network without sacrificing coverage, capacity or customer satisfaction. In fact, all three of these things improve markedly with picocells.
-9-
About ip.access
Based in Cambridge, UK, ip.access ltd. is a leading manufacturer of cost-effective picocell and femtocell infrastructure solutions for GSM, GPRS, EDGE and 3G. These solutions bring IP and cellular technologies together drive down costs and increase coverage and capacity of mobile networks. ip.access is the company behind the multi-award winning Oyster 3G™ femtocell, which dramatically improves the 3G experience at home. The ip.access nanoGSM® is the world’s most deployed picocell solution, carrying over a billion minutes of voice traffic every year for operators all over the world, including T-Mobile, TeliaSonera, and Telefónica O2. It provides GSM, GPRS and EDGE coverage and capacity for offices, shops and using satellite backhaul passenger aircraft, ships and in remote rural areas. With deployments in more than 40 live networks around the world and growing, ip.access is the partner of choice for operators competing in the new converged marketplace. ip.access is an active member of the Femto Forum, ETSI and the Network Vendors Interoperability Testing Forum.
[email protected]
Copyright © ip.access 2009. Oyster 3G™, nanoGSM® and nano3G™ are trademarks of ip.access ltd. All other trademarks are acknowledged. This document contains advance information, subject to change without notice. No responsibility is assumed by ip.access for the use of this information, nor for infringements of patents or other rights of third parties. This document is the property of ip.access and implies no license under patents, copyrights or trade secrets. No part of this publication may be copied, reproduced, stored in a retrieval system, or transmitted, in any form of any means, electronic, photographic, or otherwise, or used as the basis for manufacture or sale of any items without the prior written consent of ip.access.
- 10 -
ip.access ltd Building 2020 ■ Cambourne Business Park ■ Cambourne ■ Cambridge ■ CB23 6DW ■ UK T +44(0)1954 713700 ■ F +44(0)1954 713799 ■ [email protected] www.ipaccess.com
- 11 -
