Top ten challenges to Femtocell deployment
By Manish Singh, Continuous Computing
Femtocell is the next "big thing" in the wireless industry. Recent developments lend weight and credibility to this fledgling technology: ambitious partnerships are forming, major field tests and operator trials are underway, and industry heavyweights are joining the Femto Forum--the independent industry association that supports femtocell deployment worldwide--in record numbers. Membership has jumped from seven companies to almost sixty in less than eight months. The femtocell market opportunity, estimated to be as high as $22.5 billion by 2013, has caught the attention of incumbents and start-ups alike, and is now capturing a significant amount of analyst and media attention as well.
The question now is how quickly femtocell equipment manufacturers and network operators can overcome significant challenges, both technical and commercial, on their way to making femtocell the technology of choice for in-home wireless use. Time-to-market is absolutely critical because WiFi and Unlicensed Mobile Access (UMA) have staked out an early lead in the race for technological dominance. In this article, we offer the top ten challenges to femtocell deployment.
1. Low-cost Implementation
The cost of femtocell equipment is of paramount concern to most industry watchers. The technology simply will not take off unless femtocells are affordable to the mass market. Many vendors say they plan to charge about $200 per femtocell unit, which is too high for mass adoption. For a viable deployment, a selling price well below $200 with volume is required. In the meantime, carriers will be forced to subsidize the units heavily in order to keep prices low and encourage early adopters. In the first U.S. femtocell trial, Sprint sells its Airave femtocell units for about $50 while charging customers a flat monthly rate of $15 for individual plans and $30 for family plans. This offering matches up well with T-Mobile's HotSpot@Home.com service, which is priced at $19.99 per month and requires the purchase of a router for about $50. Since subsidizing femtocell equipment costs is not a viable long-term strategy, the industry must figure out how to lower the cost of femtocell customer premise equipment (CPE).
A key driver supporting femtocell cost reduction is integration of software with silicon--so-called "femtocell-on-a-chip." Such integration lowers development cost by building in performance features and functionality, thereby reducing the number of components needed to build a femtocell. In essence, lower chip count equals lower cost. Integrating femtocell technology into set-top boxes or cable/digital subscriber line (DSL) modems will provide further opportunities for cost reduction as these converged CPE devices provide synergies which mobile operators can exploit. In addition, converged CPE devices share certain processing and other functions, and thereby drive bill of materials (BOM) cost down through component re-use. Significant service differentiation can be achieved through converged devices, providing a path to meeting volume objectives. The opportunity for converged CPE vendors is to develop converged devices that deliver true "Quad-Play" capability cost effectively--voice, video, data, and mobility.
Low cost femtocell implementation also depends on scalability. If enough users sign up for services, operators can avoid the cost of building and maintaining expensive macrocells. Large-scale adoption is necessary in order for cost savings to be achieved. A typical small operator can save an average of about $45 per customer per year by deploying 3G femtocells in 60 percent of customer households by 2012. But if only 20 percent of an operator's customer base signs up for femtocells by that time, then the operator would save only $20 per customer per year because significant numbers of macrocells would still be needed. So in order for femtocell investments to pay off, operators must roll-out service rapidly to a large number of customers--femtocell equipment must not only be affordable but readily available. As the Femto Forum explains when identifying one of the main challenges with femtocell technology, "Operators need to be confident they can source devices in the right volumes and at the right price point."
2. Network Architecture Harmonization
Another challenge for femtocell deployment is the absence of a single, industry-standard architecture for integrating femtocells into mobile core networks. The growing interest in femto technology has fostered an environment in which there are many different, and sometimes proprietary, integration methods. At least 15 different femtocell architectures have been identified, spanning all air interfaces including CDMA, GSM, WCDMA, and WiMAX. Such diversity will translate into market fragmentation, which will in turn jeopardize scalability, complicate interoperability, and ultimately increase roll-out costs. As one analyst has noted, "[T]he introduction of open standards will be critical in enabling an economy of scale that will better enable the OEM and semiconductor communities to meet very aggressive price points and stimulate the market."
In response to the growing need for harmonization, the Femto Forum has kicked off an initiative designed to harmonize femtocell network architecture and set the stage for the development of future standards. There are two broad categories of network architectures: UMTS-centric and SIP-based. The former leverages the mobile core for hand-off functions but does not offload the mobile core. Conversely, SIP-based architectures offload the mobile core by delivering the traffic either directly to the IMS core or to a mobile softswitch-based implementation. In this case, more functionality is included within the femtocell, and integration to the core is done using an IP-based technology. New applications like Voice Call Continuity (VCC) are required to handle seamless and secure hand-offs between femtocell and macrocells.
At the Femto Forum's plenary meeting in March 2008, the femtocell community will "come together to forge a future technology path through consensus," says Simon Saunders, Chairman. "This conjoined approach will not only encourage interoperability and increase economies of scale thereby helping keep costs low, but it will also help to support far-reaching new femtocell applications."
3. Remote Device Management and Software Upgrades
Once femtocells are deployed in mass volumes, remote software upgradeability will be vital to address operators' architectural evolution in a cost-effective manner. As picoChip, a leading provider of femtocell silicon, points out, "[H]aving the ability to provide software upgrades is becoming increasingly important as standards evolve and enhancements are made." For each femtocell deployment network architecture the software architecture inside the femtocell is different, as is depicted in the following diagram:
In other words, the software residing in a femtocell changes depending on the deployment architecture, yet the underlying hardware architecture remains the same. In order to avoid "truck rolls" and maintain customer satisfaction, operators will need to be able to manage femtocells remotely. They must be able to update software and firmware, monitor the status and performance of the femtocell, and do diagnostic tests, all from a console in the operator's network. To this end, femtocell developers are working to ensure that femtocell CPE can be easily and securely configured, diagnosed, and managed remotely. This has led to increasing interest in TR-69 for femtocells, which was standardized by the DSL Forum. TR-69 will need to be extended and new profiles for femtocells need to be developed.
4. RF Interference
One of the key issues to be addressed for successful femtocell deployment is minimizing radio frequency (RF) interference, as femtocells work in licensed spectrum. If a single frequency CDMA system is being operated, where the macrocell and femtocell network utilize the same frequency band, then the power control algorithms of the femtocell can create interference with the macrocell, thereby degrading network capacity and quality of service. In multi-dwelling units such as condominiums and apartments, multiple femtocells can interfere not only with the macrocell network, but also with each other.
Some carriers are planning to deploy femtocells on a different RF frequency than their macro cellular network and thus avoid interference, but spectrum acquisition is costly. Another possible solution is to utilize the mode-2 fixed power option available in the 3G configuration parameters, which would prevent the mobile unit power from increasing and causing interference, although there could be a performance trade-off. Vendors are exploring auto-configuration as a way to address interference issues. Femtocell manufacturers are working to develop sophisticated algorithms--in essence, smart radios--for femtocells that will mitigate potential interference and adjust signals based on their environments. Although modeling by carriers shows that these algorithms could address the interference problem, they pose certain operational challenges. For example, a necessary input is the macrocell transmit power, which would require the carrier to configure the femtocells centrally. Another required known value is the line-of-sight distance to the femtocell, which would require the carrier to know where the femtocell is installed.
5. Potential Consumer Concerns
Worries persist about the safe use of radio waves and wireless communication equipment. Consumers may perceive greater safety risk from femtocells, which are often referred to as "home base stations," especially when children are present in close proximity. As one commentator points out, "[T]here is a danger that if people start hearing about 'home base stations' there will be a backlash before progress is even made."
The Femto Forum has already moved to address potential health and safety concerns in a paper entitled Femtocells and Health, stating that femtocells must comply with the same safety limits that are applied to other wireless devices such as mobile phones and their antenna sites, and concluding that, "There are no established health effects from exposure to radio waves below the limits applicable to wireless communications systems." The wireless industry will need to do more to educate the public as to why femtocells do not present a health hazard.
Access control could also be a concern for consumers, and operators must be prepared to address these questions. The prospect of allowing passers-by or neighbors in multi-tenant buildings to "free ride" on their home femtocell likely will not sit well with many consumers, especially because the consumer is paying for the broadband Internet connection being used for backhaul. Operators will need to develop access control mechanisms enabling the femtocell owner to easily add and delete authorized users and to exclude unauthorized users.
Consumer concerns about security must also be allayed. Because femtocells use residential IP broadband connections, consumers will be plugging femtocells into the Internet via cable modems and DSL routers. Use of the Internet as a backhaul network raises a number of security issues: how to protect the integrity of the operators' core networks from the public environment of the Internet, how to protect the integrity of users' traffic and how to support seamless transitions between the macro and femtocell networks. As such, each femtocell will have to establish an encrypted IPSec tunnel. To deter hacking, a femtocell must identify and authenticate itself to the operator's network as being valid. Security and mobility gateways will be needed to handle very large numbers of IPSec encrypted tunnels.
In this article we discussed five key femtocell challenges that need to be addressed and their potential solutions. In Part II we will discuss additional challenges including Quality of Service and Traffic Prioritization, Timing and Network Synchronization, Provisioning, Regulatory Hurdles and Marketing. Challenges abound, yet it is exciting to see the telecom ecosystem coming together to develop compelling solutions to address the emerging femtocell opportunity.
6. Quality of Service and Traffic Prioritization
Due to a stronger RF signal based on better indoor coverage, femtocells will provide faster handset data speeds than what mobile subscribers may be used to, which is expected to encourage greater use of data services and video streaming inside the home. This is probably the most important strategic reason for a mobile operator to deploy femtocells in its network, as femtocells could enable the mobile phone to compete not only with the fixed line phone, but also with the PC and TV for entertainment and information services in the home. Voice is real-time and latency-sensitive, which means that best-effort IP delivery is not good enough. Delivering high quality of service is a major challenge when there is a shared IP access link simultaneously carrying voice, video, peer-to-peer (P2P) traffic, and data. In this operating environment, traffic prioritization becomes essential--not only for effective handling of real-time voice packets, but also for prioritizing network time protocol (NTP) or IEEE 1588 timing packets. If timing packets are not prioritized, then femtocells can go out-of-sync and lead to dropped calls, which results in poor subscriber satisfaction. DSL uplink bandwidth, in certain deployments, adds additional constraints. In the U.S., for example, depending on the distance from the CO, many mid-tier plans have only 200K throughput on uplink. With other devices like PCs that have active VPN connections sharing the same uplink with the femtocell equipment, traffic prioritization becomes essential to ensure consumer experience does not deteriorate with events like VPN connection breaking when making a mobile voice call.
7. Timing and Network Synchronization
Attaining required timing and network synchronization levels is a major femtocell challenge. 3GPP specifies that base station transmit frequencies be very accurate and closely synchronized, requiring precise clock references which add to base station costs. The standard has been relaxed for picocells in Release 6, and it has been proposed that future versions of the standard relax it even further. Lower-cost synchronization solutions are still needed.
To meet stringent network synchronization requirements, femtocells can use IEEE 1588 for an accurate clock synchronization protocol between networked equipment. As picoChip points out, "[B]y distributing a high precision time base around the network that is resilient to the typical levels of packet delay and jitter found on broadband Internet connections, a low-cost implementation is possible." Alternatively, a GPS timing reference could be used, or the base station could receive transmissions from the overlaying macro cellular network and adjust its timing accordingly. Finally, there are some innovative, low-cost / high-stability temperature-controlled crystal oscillators coming onto the market that may make it easier to attain the required synchronization levels. More study is needed for this critical facet of femtocell operation. 8. Provisioning
From a consumer point of view, femtocell technology must be "plug-n-play," and hence seamless installation and activation is critically important. If the equipment is not easy to install, and if the service is not simple to activate and use, on-site support from a service technician will be required. Since "truck rolls" add to an operator's operational expenditures and run contrary to the femto value proposition, carriers must be able to achieve "zero touch installation" for successful femtocell deployment.
In order to keep femtocell customer service and technical support costs down, operators will re-use large portions of their current core networks such as billing, authentication, authorization, and accounting. Registration and provisioning could be carried out in the same way as is done by many cable and ADSL network operators. For example, in most cases, the subscriber has a password to gain access to the broadband connection. DSL Forum's TR-069 standard that already exists for the remote provisioning and management of DSL gateways could be extended for the provisioning of femtocells. Alternatively, the femtocell could include a Subscriber Identity Module (SIM) that stores a key to identify the user. When the femtocell is connected and switched on, the SIM is used to authenticate with the network and provide a secure connection.
Either way, operators must closely manage the customer experience in order to ensure a smooth and swift installation process. The need for standardizing remote device provisioning and activation is a must to ensure that carriers can deploy femtocells from multiple vendors without having to change their BSS and OSS systems. It will be essential for carriers to have a strong control on the remote activation of femtocells. Carrier must be able to de-activate femtocells depending on the location or based on gray lists and black lists to prevent femtocell activation followed by reported theft.
9. Regulatory Hurdles
A number of regulatory hurdles must be overcome or removed for successful femtocell deployment. Since femtocells carry voice calls, they will be required to support a 911 emergency service. Alternative power sources or a fall-back to existing telephone infrastructure may be viable ways to achieve E911 availability. To meet E911 requirements, the operator must be able to provide the location of the equipment to the Public Safety Answering Point (PSAP), which poses a challenge because consumers might try to carry their femtocells to new locations. Some manufacturers are using GPS to lock the equipment when it is moved to a different location. In some countries, existing licenses require operators to maintain written records of base station locations. The need to create and maintain records of every femtocell is unnecessary and would create an excessive burden on operators. Such records are not required for existing systems such as cordless phones and wireless LAN access points, so many operators argue that femtocells should also be exempt from this registration requirement.
In some countries such as Japan, a licensed technician is required to install a "base station." If femtocells are classified as base stations by regulatory bodies and are not exempted from the licensed technician installation requirement, this regulation will defeat the femtocell value proposition because network rollout cost will significantly increase. To anticipate and proactively respond to these and other potential regulatory hurdles, the Femto Forum--the independent industry association that supports femtocell deployment worldwide--is already working with its members and international regulators to encourage a benign regulatory environment for femtocell deployments.
10. Marketing
Initially, operators have focused on the technical challenge of integrating femtocells into their mobile networks, but they must not overlook the importance of successfully positioning femtocells within the industry and amongst the wider public. In addition to developing a common terminology and branding the service effectively, it is critical to build robust usage cases and marketing plans to drive femtocell adoption. To be successful, operators will need to develop compelling value propositions aimed at the most appropriate segments. Enhanced indoor coverage and reduced call charges for voice calls made in the home may not be enough to persuade mobile users to adopt femtocell technology, especially in light of alternatives based on UMA / WiFi.
3G femtocells could clearly boost revenues by allowing mobile operators to offer relatively inexpensive, high-quality voice calls for mobile users when they are at home while charging a premium for mobile calls when users are elsewhere. Operators could also enjoy certain "soft" benefits such as increased uptake and reduced churn, by offering various group subscriptions and family plans linked to femtocells in order to attract all the members of a household to their networks. Furthermore, femtocells will enable mobile operators to compete against converged cellular/WLAN services by offering very similar plans without the need for special handsets. Regardless of the tactics, for femtocells to be successful, the industry will need to have a smart, targeted, and differentiated way to market the technology.
Final Note
The trends are clear: voice traffic will continue to migrate from fixed to mobile networks until 3G carries the majority of voice telephony. Likewise, consumer demand for mobile data in the home is growing and has significant future growth ahead. In the battle for the home, first-mover advantage is key, and femtocell technology has great potential but the race with WiFi is definitely on. In this "make-or-break" competitive environment, fast time-to-market could make the difference between wildly successful deployment of femtocell--leading to broad societal transformation " and another promising technology relegated to the dust-bin of history. With the help of many dedicated and talented telecom ecosystem professionals in meeting these challenges, femtocells will live up to the promise.
