Small cells challenge the macronetwork model

By Julian Bright

The fragile economic climate and continued downward pressure on operator spending probably helped ensure that this year’s Mobile World Congress in Barcelona was a more grounded and businesslike affair than has been the case in some previous years. It’s clear that the explosion in mobile data traffic, coupled with the squeeze on operators’ capex and opex, is prompting some radical thinking about the future shape and structure of the network. Increasing network capacity will be critical, but until operators start to see data traffic turning into revenues, they need to focus on optimizing network performance and lowering costs.

The traditional macronetwork model is under fire as operators look for more cost-effective deployment strategies that target network capacity where it is most urgently needed. It is a trend that doesn’t end with the RAN but extends into the backhaul and core networks and is coupled with the drive to reduce the per-bit cost of transporting data across the entire network.

Network vendors can be relied on to identify an opportunity whatever the underlying state of the market, and at MWC they were eager to engage in a dialogue about their plans to create capacity and lower costs. Offloading a proportion of data traffic to free up valuable network capacity was a dominant theme of the week, with femtocell vendors, Wi-Fi providers and any other flavor of alternative network-capacity provider eager to enter the debate.

In a display of magnanimity rarely seen among a group of vendors, however, there was general acknowledgement that no single technology holds the answer to the problem and that coexistence of a number of technologies would be the only way forward.

Even so, the “femtocell vs. wifi” debate continued to be played out, hinging on a range of commercial and technical issues. Femtocell providers stressed the benefits to operators of keeping customers “on-net” while achieving cost savings and capacity gains in the macronetwork. Meanwhile, the bill of materials for femtocells is expected to be quickly driven down to below $100 as new manufacturers, such as wifi-access-point and DSL-modem suppliers, enter the femtocell market.

The vendors pointed to concerns surrounding Wi-Fi, such as device battery life and the likelihood that most midrange feature phones would not be wifi-enabled. wifi providers responded by saying that femtocells fail to target the real network hot spots and that the lack of planning behind femtocell deployments could lead to interference problems, particularly in areas of dense deployments.

The interference concerns surrounding femtocells were echoed by picocell providers and providers of alternative in-building distributed antenna systems, which also argued that femtocells didn’t scale well in larger implementations, where traffic engineering was key. But femtocell vendors say that incorporating self-organizing-network (SON) capabilities into their products will overcome many of these interference concerns.

Femtocell vendors say their technology is starting to gain traction, with a dozen operators having either introduced commercial services or made firm commitments to do so, and withy offerings such as Vodafone’s Sure Signal gaining market acceptance. And although the initial business case is built largely on coverage and capacity, femtocell-based services and a richer application environment will inevitably follow, they say.

New architectures

As expected, LTE was a major theme at MWC, but with the emphasis more clearly on optimizing the network to lower costs. The announcements by US carriers Verizon and AT&T in 2009 of their LTE-deployment plans, coupled with the early soft launch by TeliaSonera in Sweden and Norway late last year, have gone some way toward dissipating the discussion about LTE launch dates.

However, there is growing concern that the explosion in data traffic from devices such as the iPhone is producing little in terms of additional revenue for operators, and LTE is increasingly regarded as the way for operators to optimize their infrastructure and thus drive down costs.

The consensus at MWC appeared to be that a new network architecture composed of smaller cell sites as a complement to the macronetwork rollout would be the most cost-effective means of targeting high-usage areas. Microcells, picocells and femtocells – this time in the form of outdoor “metro femtos” – represented one option, possibly even as part of an “inside out” network-rollout strategy whereby LTE would be deployed in traffic hot spots in advance of a macronetwork rollout.

Japan’s NEC maintains that running a macronetwork at the 2.6GHz frequency likely to be used for LTE rollouts in Europe doesn’t make sense and that deploying LTE in small cells where there is actual capacity demand is potentially simpler, cheaper and faster than a macro deployment.

The big worry for operators, according to Motorola, is that they will invest in LTE infrastructure but that content vendors will make most of the money out of it. With the cost of the base station representing an increasingly small portion of the overall investment in next-generation mobile networks, vendors such as NEC, Motorola and Alcatel-Lucent are focusing on end-to-end offerings that provide convergence in the radio-access, backhaul and core networks.

As expected, voice services for LTE were also under discussion in Barcelona, with the GSMA announcing that it is driving work on the One Voice IMS-based initiative, which has the backing of more than 40 companies and organizations. Nevertheless, there is some skepticism regarding the performance of CS Fallback, the interim technology recently endorsed by the Next-Generation Mobile Networks alliance, with at least one leading operator, T-Mobile, claiming that VoLGA is the superior technology.

The word at MWC, however, was that US operator Verizon intends to move directly to a One Voice approach, with LTE handsets in the market as early as 1Q11.