Leading operators like Vodafone and Orange have been insisting that the femtocell industry standardises on the Iu interface, which is the existing common reference point between the core network and the radio access network. It is frequently used by operators to allow them to purchase radio networks from multiple, different equipment vendors and connect them to their core networks. This gives them greater purchasing power, and be able to trade-off between multiple vendors.
In the 3G standards, there is also an Iub interface, which was originally defined by operators to allow them to buy their 3G basestations from different suppliers than the 3G Radio Network Controller (RNC). In practice, however, the equipment costs for 3G radio networks was driven down, and the complexity and risk of managing multiple vendors within the same radio access network so high, that this has never taken off.
Femtocell vendors have been strongly encouraged to adopt the Iu interface, so that their equipment can be interconnected with the standard 3G core network with minimum change. This allows the full set of existing services to be provided to the end user.
There are several alternative subsystems promoted both above and below the Iu interface as shown below:
By far the most likely scenarios, at least in the short/medium term are based on reusing the existing 3G core network above the Iu interface (so not SIP or IMS) connected to either the UMA standards based or proprietary subsystems. Since both the UMA standards based solution and the proprietary Iu based alternatives comply with Iu and can use existing standard 3G phones, the major difference is more about vendor choice. Whilst it might in theory be possible to use multiple vendors in a UMA architecture, there is really only one UMA gateway vendor at present - other vendors have UMA gateways for WiFi but haven't yet announced 3G femto versions. Other vendors, including Motorola and NEC, have incorporated the Kineto Wireless UNC capability into their overall solution.
Several vendors propose SIP based solutions (SIP is the IP protocol used to make and receive phone calls over the internet). We don't see any of these being mainstream options at present:
- Sonus networks have an existing IP core network, which handles billions of minutes a month for the long distance carriers (mainly in the US). Their core network is SIP based and can be used via access gateways to provide calls and switchboard type services within a business (sometimes called Centrex). They have developed a gateway which converts from the Iu interface into their core network, and thus could provide the same services across both fixed and mobile phones. Since the femtocell must be provided and managed by an operator with licenced spectrum, we see this option as a corner case - the phone wouldn't work when in the outdoor macrocellular phone environment, and roaming wouldn't be easily supported.
- Tatara have proposed a SIP based solution for use within the radio access network, effectively communicating between the femtocell and gateway controller using the SIP protocol. However, existing 3G phones would use their existing protocols, and the Iu interface would also be complied with. Thus all this would achieve is a conversion of protocols between the handset into SIP and back out again into Iu for little or no advantage. Whilst this might be considered futureproof, because SIP might be adopted some years ahead, we also cannot see the benefit of this.
- Standalone femtocells are effectively a complete mobile network in a box, and communicate with the outside world using SIP (to make voice calls) or data, are quite feasible by adding more software into the box. However, operators are likely to be concerned about being bypassed completely and losing control of their customers and traffic. As licenced owners of the spectrum, they will want to be able to manage and update the configuration of any equipment that could interfere with their macrocellular network.
Within the standard radio network, there are really two options, both of which work with standard 3G phone and provide a standard Iu interface. The choice is whether to conform to the UMA standard or not:
- UMA vendors, which centre around Kineto Wireless (the only significant UMA gateway vendor) and Ubiquisys (which provides femtocell technology to Netgear and possibly other vendors). Technically, it should be possible for an operator to connect multiple different vendor's femtocells to the same UMA gateway and thus give the operator more choice. NEC has acted as a systems integrator for a Kineto/Ubiquisys badged solution, which is being trialled by several operators. Motorola are also believed to incorporateKineto UMA technology and Ubiquisys femtocells combined with their own Netopia device management product.
- Co-operating vendors working around their own proprietary standards. The leading vendor here is Nokia-Siemens, which has decided not to build their own femtocell, but instead provide a femtocell gateway and publish their own interface. Several of the independent femtocell vendors have announced interworking with this proprietary protocol.
- Radio Access Network vendors, such as Alcatel-Lucent and Huawei, have chosen to provide a complete end to end solution with their own equipment. They can use their own proprietary protocols within their subsystems, and standardise only at the edges of the solution.
So what's the most likely outcome? We'd expect the operators to insist on the Iu interface and to continue to use their existing core networks - SIP and IMS based solutions are many years away. UMA has the moral advantage by being only the standardised solution for 3G femtocells, and allows operators to buy femtocells from different vendors whilst interworking with the same UMA gateway controller. An open standard interface for the cellsites has not been signficant for macrocellular 3G networks - typically operators have one or two suppliers and separate them geographically to avoid interworking problems. Operators may demand more interoperability between femtocell vendors, so that they can mix and match supplies without major changes to their supporting gateways and management systems. Others may simply buy complete solutions comprising both gateways and femtocells from the same vendor. Thus femtocell vendors with a proprietary interface between femto and gateway controller are not at a disadvantage, at least in the short term. Perhaps when there are many millions of units deployed, the operators may think again.
Device Management Standards
Most (if not all) femtocell vendors have adopted the existing TR-069 device management standard created by the DSL forum. This is widely deployed in the DSL broadband industry, and there are several manufacturers of access controllers. By using the standard, existing DSL management practices can be extended to handle the femtocells, whether they are integrated into a DSL modem or operate standalone. Leading vendors are Motive (who would claim to have authored much of the original standard) and Netopia (now owned by Motorola), but there are many competitors in this space.