Femtocell

From Wikipedia, the free encyclopedia

Jump to: navigation, search

In telecommunications, a femtocell—originally known as an Access Point Base Station—is a small cellular base station, typically designed for use in residential or small business environments. It connects to the service provider’s network via broadband (such as DSL or cable); current designs typically support 2 to 4 active mobile phones in a residential setting. A femtocell allows service providers to extend service coverage indoors, especially where access would otherwise be limited or unavailable. The femtocell incorporates the functionality of a typical base station but extends it to allow a simpler, self contained deployment; an example is a UMTS femtocell containing a Node B, RNC and GPRS Support Node (SGSN) with Ethernet for backhaul. Although much attention is focused on UMTS, the concept is applicable to all standards, including GSM, CDMA2000, TD-SCDMA and WiMAX solutions.

For a mobile operator, the attractions of a femtocell are improvements to both coverage and capacity, especially indoors. There may also be opportunity for new services and reduced cost. The cellular operator also benefits from the improved capacity and coverage but also can reduce both capital expenditure and operating expense.

Femtocells are an alternative way to deliver the benefits of Fixed Mobile Convergence. The distinction is that most FMC architectures require a new (dual-mode) handset which works with existing home/enterprise Wi-Fi access points, while a femtocell-based deployment will work with existing handsets but requires installation of a new access point.

Contents

[edit] History

The concept of a compact self-optimising home cellsite has been documented since 1999. Alcatel announced in March 1999 that they would bring to market a GSM home basestation which would be compatible with existing standard GSM phones.The high unit cost made the product unviable.[1]

Various research projects continued to work on femtocell concept products, with Motorola engineers in Swindon designing the first complete 3G home base station in 2002 termed an 'Access Point Base Station' which had no requirement for a core network, mimicing a WLAN access point but for 3G and WiMAX. A derivative was later marketed as the Motorola AXPT. [2]

Several alternative technical solutions were developed, such as UMA which uses dual-mode WiFi/cellular handsets to achieve the same goals.

By mid-2004 a number of companies were independently investigating femtocells (although mostly using other terms such as "residential base station" or "3G access point").[3]

Also in 2004, two femtocell-focused companies were registered[4] at Companies House in the United Kingdom: 3Way Networks (now part of Airvana, Inc.) and Ubiquisys. By 2005, the idea had become more widely recognised with demonstrations and conference discussion. By this stage more companies were involved, including more established suppliers Samsung, Airwalk, ip.access and RadioFrame Networks.

By early 2007, the idea had become mainstream, with a number of major companies publicly demonstrating systems at the cellular industry 3GSM conference in February, and operators announcing trials. In July, the Femto Forum trade organisation was founded to promote femtocell deployment worldwide,[citation needed] comprising mobile operators, telecoms hardware and software vendors, content providers and start-ups. Its main work is conducted via four working groups, tackling regulatory issues, network and interoperability, radio and physical layer, and marketing and promotion.

In 3Q 2007, Sprint Nextel started a limited rollout in Denver, Indianapolis and Tennessee of a home-based UbiCell femtocell built by Samsung called the "Sprint Airave", which worked with any Sprint handset.[5] Airave was rolled out nationwide on 17 August 2008.[6]

As well as system manufacturers, semiconductor companies have announced chip-level products to address this application. Analog Devices has developed a chipset for the RF-IF and baseband, while picoChip claims significant commercial traction on their baseband Digital Signal Processor[citation needed]. There are significant number software stack providers for the femtocell based base stations. Continuous Computing [7] has announced complete solution for femtocell based software stacks adhering to various femtocell based UMTS architecture described below.

[edit] Issues

Although claims are made that Femtocells could be a panacea for straightforward system deployment, there are a number of complications that need to be overcome.

[edit] Interference

The placement of a femtocell has a critical effect on the performance of the wider network, and this is one of the key issues to be addressed for successful deployment.

Without unique spectrum for the femtocell 'underlay network', or very careful spectrum planning in the wider network, there is a concern that femtocells could suffer from severe interference problems. For example, in a femtocell handover between a macrocell network to a home femtocell access point, there are limitations in the standards which must be taken into account. For example, there is a limitation in the number of adjacent cell sites - typically 16 - for which the mobile unit can scan for, measure and then pass to the RAN handover algorithm (for 2G and 3G standards, for example). Further, if a single frequency CDMA system is being operated, where the macro and femtocell network utilise the same frequency band (a typical situation for many operators who licensed only one 3G frequency band), then the power control algorithms of the macro cell and femtocell can create interference ,[8] where for example a mobile unit increases its transmit power to the femtocell as part of the 'near-far' power control inherent in CDMA systems, whilst it is within the coverage area of a macro unit. The resultant high power transmitter in the macro field acts as an interferer since the frequency is shared. Finally, there is the issue of coverage area, where in high-rise accommodation, femtocell users on different floors can create interference to other users. There are several partial solutions to this problem, but primarily the only way to prevent interference is to use a different frequency for the femtocell coverage, particularly for CDMA deployments. The partial solutions include utilising the mode-2 fixed power option available in the 3G configuration parameters, which would prevent the mobile unit power from increasing and causing interference, though there is an obvious performance trade-off if this approach is used.

Many vendors are reported to have developed sophisticated algorithms to address the problem, and modelling by carriers indicates this is viable.[citation needed] As such, the trials now in place are designed to test these techniques and to determine to what degree interference is a problem and under what circumstances. In his paper for 'PIMRC 07',[9] Claussen describes the UMTS femtocell/macrocell interference problem and concludes that to manage the interference that "Essential requirements such as autoconfiguration and public access" are needed. In this case 'public access' means that all deployed femtocells using the same frequency (ie. of the same operator) would need to allow anyone to access the femtocell; there are obvious backhaul issues with this if the user is paying for the DSL or Cable backhaul connection. It is suggested in the paper that this could be offset by low cost calls. In another paper,[10] Ho and Claussen identify the pre-requisite for auto-configuration of the femtocell power level in order to reduce interference - though in Claussen's first paper the algorithm requires knowledge of the macrocell transmit power, which would require the operator to configure the femtocells centrally, and line-of-sight distance to the femtocell, which requires knowledge of where the femtocell is installed. In his second paper, Ho highlights the issue of increased network traffic due to handover messages between the macrocell and femtocell.

The 3GPP meeting reported that: "To the extent investigated so far co-channel deployment is feasible for open access. For closed access, analysis conducted so far indicates that co-channel deployment is feasible if adaptive interference mitigation techniques are used. Further work is required to summarise the trade-off between HNB performance and the impact on the macro layer and to determine whether an acceptable tradeoff can be identified".[11]

A number of companies [12] are using the approach of using the femtocell as a mobile phone (UE) in order to measure, synchronise and build a neighbour list of nearby base stations. From this information, power levels, spreading codes and other parameters can be determined and resolved in order to avoid interfering with existing infrastructure.

[edit] Spectrum

Crucially, access point base-stations operate in licensed spectrum. As licensed spectrum allocation is made to operators on a fee basis, deployment of equipment must meet the strict requirements of the licenses. To make best use of spectrum, operators use frequency and cellular planning tools to optimise the best coverage for a given amount of spectrum. The introduction of access point base stations using licensed spectrum that are sold directly to the customer has implications for frequency and cellular planning, since an unexpectedly located access point base station could interfere with other closely-located base stations.

[edit] Access control

There is also the related issue of what happens when a neighbor's mobile appliance attaches to the network using another neighbor's femtocell, or how that can be prevented from occurring.

[edit] Lawful interception

Access point base stations, in common with all other public communications systems, are, in most countries, required to comply with lawful interception requirements.

[edit] Equipment location

Other regulatory issues[13] relate to the requirement in most countries for the operator of a network to be able to show exactly where each base-station is located, and for E911 requirements to provide the registered location of the equipment to the emergency services. There are issues in this regard for access point base stations sold to consumers for home installation, for example. Further, a consumer might try to carry their base station with them to a country where it is not licensed. Some manufacturers (see Ubicell) are using GPS within the equipment to lock the femtocell when it is moved to a different country;[14] this approach is disputed, as GPS is often unable to obtain position namely indoors because of weak signal.

[edit] Network integration

From an operational or deployment perspective, one of the key areas that needs to be considered is that of network integration. A conventional cellular network is designed to support a relatively small number (thousands, tens-of-thousands) of base stations, whereas a femtocell deployment of millions of consumer access points requires a different architecture to support this scaling. The issue of increase in network traffic as a result of co-channel macrocell / femtocell deployment is discussed in the paper by Ho and Claussen.[15]

[edit] Emergency calls

Access Point Base Stations are also required, since carrying voice calls, to provide a 911 (or 999, or 112) emergency service, as is the case for VoIP phone providers.[16] This service must meet the same requirements for availability as current wired telephone systems. There are several ways to achieve this, such as alternative power sources or fall-back to existing telephone infrastructure.

[edit] Quality of service

When utilising an Ethernet or ADSL home backhaul connection, an Access Point Base Station must either share the backhaul bandwidth with other services, such as Internet Browsing, Gaming Consoles, set-top boxes and triple-play equipment in general, or alternatively directly replace these functions within an integrated unit. In shared-bandwidth approaches, which are the majority of designs currently being developed, the effect on QoS may be an issue.

The uptake of femtocell services will depend on the reliability and quality of both the cellular operator’s network and the third-party broadband connection. When things go wrong, subscribers will turn to cellular operators for support even if the root cause of the problem lies with the broadband connection to the home or workplace. Hence, the effects of any third-party ISP broadband network issues or traffic management policies need to be very closely monitored and the ramifications quickly communicated to subscribers.

A key issue recently identified being active Traffic shaping by many ISPs on the underlying transport protocol IPSec. UK-based femtocell authority Epitiro have recently provided significant publicly available research and insight into many of these IP-focused QoS issues. A femtocell deployment guide from Epitiro is available for download here.

[edit] Spectrum accuracy

To meet FCC/RA spectrum mask requirements, Access Point Base Stations must generate the RF signal with a high degree of precision, typically around 50 parts-per-billion (ppb) or better. To do this over a long period of time is a major technical challenge, since meeting this accuracy over a period longer than perhaps 12 months requires an ovenised crystal oscillator (OCXO). These oscillators are generally large and expensive, and still require calibration in the 12-to-24 month time frame. Use of lower-cost temperature-compensated oscillators (TCXO) provides accuracy over only a 6-to-18 month time frame. Both depend on a number of factors.

The solutions to this problem of maintaining accuracy are either to make the units disposable/replaceable after an 18-month period and thus keep the cost of the system low, or to use an external, accurate signal to constantly calibrate the oscillator to ensure it maintains its accuracy. This is not simple (broadband backhaul introduces issues of network jitter/wander and recovered clock accuracy), but technologies such as the IEEE 1588 time synchronisation standard may address the issue, potentially providing 100-nanosecond accuracy (standard deviation),[17] depending on the location of the master clock. Also, Network Time Protocol (NTP) is being pursued by some developers as a possible solution to provide frequency stability. Conventional (macrocell) base stations often use GPS timing for synchronization and this could be used to calibrate the oscillator.[18] However, for a domestic femtocell, There are concerns on cost and the difficulty of ensuring good GPS coverage.

Standards bodies have recognized the challenge of this and the implications on device cost. For example, 3GPP has relaxed the 50ppb precision to 100ppb for indoor base stations in Release 6 and has proposed a further loosening to 250ppb for "Home NodeB" in Release 8.

[edit] Handover

In order to ensure that the user gets the best data rate out of the system, the mobile appliance must somehow know to connect to the femtocell when within range, even if there is still sufficient signal from, for example, an external macrocell base station. Forcing the mobile appliance to do this, whilst preventing your neighbor's mobile appliance from doing the same, is quite a challenge. In addition, handoff from the femtocell to the wider area macrocell and back again is potentially quite complex.

[edit] Air Interfaces

Although much of the commercial focus seems to have been on UMTS, the concept is equally applicable to all air-interfaces. Indeed, the first commercial deployment is the cdma2000 Airave.[19] Femtocells are also under development for GSM, TD-SCDMA, WiMAX and LTE. The LTE study groups have identified femtocells ("Home eNode B") as a priority area.

[edit] Architectures

[edit] Home Node B (HNB)

In May 2008, the 3GPP completed a feasibility study of femtocell network architectures. Architectures including Cellular Base Station, Collapsed Stack and UMA/GAN were evaluated. As a result, the 3GPP is pursuing a new Home Node B (or HNB) reference architecture which builds on elements from both the Collapsed Stack and UMA/GAN approaches.

As the 3GPP completes the formal standard towards at the end of 2008, vendors and operators will migrate to support this new architecture for 3G femtocells.

Note the 3GPP refers to 3G femtocells as Home Node Bs (HNBs).

[edit] Cellular Base Station (Picocell)

One approach for a femtocell is to use the traditional base station architecture. In this case, the femtocell is a base station, connecting to the core network using a standard interface; for example, a WCDMA Node B connecting to a RNC via a backhaul connection (the Iub). The slight difference to a typical base station deployment is that the backhaul would be carried over broadband ("Iub over IP") which may have quality & security concerns. A more significant drawback of this architecture is that standards based base station controllers are designed to support only a limited number of high-capacity base stations, not large numbers of simple ones. This architecture was previously referred to in the literature as a picocell deployment and is one in which a base station controller is introduced to provide the necessary support to the numerous small pico-head base stations.

[edit] Collapsed Stack

More common architectures collapse some of the network functionality into the base station ("collapsed stack" or "Base Station Router"), not just the base station itself (Node B or BTS) but also the controller (e.g., RNC) and enable local radio resource control. This would then connect back to the mobile operator core at a higher point (e.g., Iu interface for WCDMA) for central authentication and management. This addresses the scalability concerns above, as the resource is located locally. The original Access Point Base Station followed this architecture but also incorporated the core MSC/GSN functions of authentication, control and switching.

[edit] Collapsed Stack with UMA Backhaul

A variant of the above is to use GAN/EGAN Unlicensed Mobile Access (UMA) standards. In this case, the UMA/GAN client is integrated into the femtocell. UMA/GAN protocol provides the connection to the mobile core, tunneling the Iu protocol. This approach uses UMA/GAN's existing security, transport and device management capabilities.

UMA/GAN is an attractive option for operators to leverage their investment in the UMA Network Controller to support applications beyond femtocells, including dual-mode handsets/WiFi or fixed line VoIP with terminal adapters.

The approach for UMA-based femtocells differs from a dual-mode handset approach where the UMA client is integrated in the device. In the former system the terminal is not affected and the air-interface is still standard - the UMA client is incorporated in the femtocell.

[edit] SIP or IMS

The final, and most sophisticated structure is to move to a full IP-based architecture. This approach was utilised in the original Access Point Base Station. In this case, even more functionality is included within the femtocell, and the integration to the core is done using an IP-based technology, e.g. SIP, IMS or H.323.

[edit] Deployment

Currently, the most significant deployment is that of Sprint. This started in 3Q/2007 as a limited rollout (Denver and Indianapolis) of a home-based femtocell built by Samsung Electronics called the Sprint Airave that works with any Sprint handset.[5] As of 17 August 2008, Airave has been rolled out on a nationwide basis.

In November 2008, Starhub rolled out its first nation-wide commercial 3G Femtocell services.

A number of operators have announced intention to have field trials in 2008, including O2,[20] Softbank,[21] TeliaSonera,[22] and Vodafone.[23]

Most analysts agree that 2008 will primarily be field trials and soft launch, while smaller, controlled commercial services may start later in 2009 [24] [25].

[edit] References

  1. ^ ThinkFemtocell - History of Femtocells
  2. ^ New 3G High Speed Indoor Access Point
  3. ^ Disruptive Wireless
  4. ^ Companies House File Numbers 05213514 and 05247998, respectively
  5. ^ a b Airave
  6. ^ Sprint AIRAVE Nationwide Launch August 17
  7. ^ Continuous Computing[1]
  8. ^ "Uplink Capacity and Interference Avoidance for Two-Tier Femtocell Networks", Vikram Chandrasekhar and Jeffrey G. Andrews
  9. ^ Performance of Macro- and co-channel femtocells in a hierarchical cell structure", Holger Claussen, Bell Laboratories Alcatel-Lucent, The 18th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications 2007 (PIMRC'07)
  10. ^ "Effects of user-deployed, co-channel femtocells on the call drop probability in a residential scenario", Lester T. W. Ho, Holger Claussen, Bell Laboratories Alcatel-Lucent, The 18th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications 2007 (PIMRC'07)
  11. ^ 3GPP TR 25.820 V1.0.0 (2007-11)
  12. ^ http://www.picochip.com/downloads/PC8209ProductBrief.pdf
  13. ^ FCC requirements for 911 provision by VoIP providers
  14. ^ Hands on with the Samsung Ubicell
  15. ^ "Effects of user-deployed, co-channel femtocells on the call drop probability in a residential scenario", Lester T. W. Ho, Holger Claussen, Bell Laboratories Alcatel-Lucent, The 18th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications 2007 (PIMRC'07)
  16. ^ FCC requirements for 911 provision by VoIP providers
  17. ^ IEEE-1588 Standard for a precision clock synchronization protocol
  18. ^ Hands on with the Samsung Ubicell
  19. ^ Sprint Customers in Select Areas of Denver and Indianapolis Get AIRAVE for Enhanced In-Home Coverage
  20. ^ O2
  21. ^ Softbank
  22. ^ TeliaSonera
  23. ^ Vodafone
  24. ^ 100,000 Femtocells Will Ship in 2008, But 2010 Will Be the Year of Real Volume, says ABI Research
  25. ^ Network World: interview with Motorola VP GM Alan Lefkof

[edit] See also

[edit] Further reading

[edit] External links

Personal tools