Fixed Mobile Convergence (FMC) provided the impetus for Unlicensed Mobile Access (UMA). UMA has been standardized by 3GPP under the Generic Access Networks (GAN) standard. The basic idea behind UMA is to enable wireless handsets to work over a wireless local area network (WLAN) such as 802.11 (WiFi) in the unlicensed spectrum range. The premise is that UMA leverages the wide proliferation of WiFi hotspots through the embedding of WiFi chipsets in mobile handsets. Once the handset is in the vicinity of a WiFi hotspot, the UMA client attaches to the hotspot and, if access is granted, the client establishes a secure IP connection to the UMA Network Controller (UNC), after which all the mobile data and voice is routed via the WLAN hotspot to the UNC. The UMA client ensures that if the user moves in and out of a hotspot, calls are seamlessly handed to and from the macrocellular network. Similar to femtocells, UMA also offloads traffic from mobile cellular networks to WiFi hotspots and their associated backhaul networks.
Femtocells have certain advantages and disadvantages when compared with UMA. For one, all the necessary components for femtocell deployment are already present in many homes and offices today: a standard mobile phone and an IP broadband connection to backhaul traffic to the operator’s network. The same generally holds true for UMA, but the main difference is that with femtocell a user need not upgrade to a new UMA-enabled handset to take advantage of better indoor coverage. Because femtocells operate in licensed spectrum, users’ existing handsets work seamlessly in the femtocell environment. Furthermore, the cost of adding WiFi functionality to a UMA-enabled handset tends to drive up its BOM cost, which is a negative factor for widespread adoption. The disadvantage of femtocells, though, is that wireless operators have to deploy a new CPE device (the femtocell) for each indoor environment, whereas with UMA the many thousands of existing WiFi connections are easily leveraged.
There are also other differences between UMA and femtocells. UMA uses unlicensed spectrum, which makes it prone to interference and can result in deteriorated voice quality. Femtocells, on the other hand, use licensed spectrum and provide standard UMTS interfaces, and thus they leverage robust channel structures and deliver good quality voice with improved coverage. What’s more, seamless handoff from the macrocellular network to femtocells and vice versa is already in place as the existing mobile core network is leveraged to deliver this function. From the consumer’s perspective, another important factor to compare is handset battery life. Because of the need for multiple chipsets, UMA handsets require more power when compared to 3G UMTS handsets and thus tend to drive battery life down. From the wireless network operator’s perspective, a key risk factor that UMA poses is cannibalization of voice ARPU because UMA-enabled handsets potentially enable competitive over-the-top voice service delivery. From an RF interference point of view, UMA operates in unlicensed spectrum and therefore does not interfere with operators’ existing macrocellular networks. On the other hand, because femtocells operate in the same UMTS licensed spectrum used in the macrocell network, dynamic Radio Resource Management algorithms are required to minimize interference with the macro-cellular network.
