There has been quite a bit of discussion lately about the need to move to small cell sizes. This is driven by the ever increasing tide of data traffic that threatens to swamp the network. In order to accommodate this a variety of strategies are being pursued: Femto cells (indoor 3G/4G access points connected to the buildings wireline connections, usually owned and operated by the building occupants), WiFi (News - Alert) offload (use of existing WiFi access points to offload data but not voice traffic) and microcell architectures (small basestations deployed at or near street level and owned and operated by service provider). While each of these solutions has its own merits and problems I am going to focus on the microcell architecture.

In general the driver to deploy microcells is to improve the user experience by bringing the base station closer to the user and reducing the number of users per base station. For the service provider this allows more total traffic to be carried over a fixed amount of Radio Access Spectrum (News - Alert), thereby increasing the total network capacity. The reasons for adopting this approach vs femto cells and/or WiFi offload are better control over the network performance and closer connection with the end customer hopefully resulting in more control of the user experience and hence more “stickiness” with the service. To address this application there have been a number of product announcements by the base station vendors with smaller radio heads, integrated radios and antennas as well as a flurry of developments taking the femtocell chips sets “upscale” to support microcell applications. There have also been a number of announcements by backhaul vendors of products targeted at this application. What seems to be missing is an integrated solution that addresses all of the requirements (power conditioning and backup, switching, environmental packaging, etc).

Also missing from most of the discussions is a comparison of the relative economics – almost as though the microcell deployment is required to deal with the network capacity and so must proceed at any cost. Of course, service providers are in business to make money and the increasing traffic requires a reduction in cost per Mbps, not an increase, in order to stay profitable. It is very important to look at the overall economics to understand if this is going to be a flash in the pan or an ongoing trend. When comparing microcells to macrocell deployments, the first part of the problem is to properly scope the analysis. In order to establish and operate a new macrocell, the total cost is measured in hundreds of thousands of dollars. Only a small portion of this actually goes to the equipment vendors and the rest goes into civil engineering, cabling, installation, power, site leasing, maintenance and support.

When compared to a microcell architecture using an integrated microcellular unit (i.e. a package that contains the base station, power, transport and switching in a single pole mountable package) we get a dramatically different picture. The majority of the civil works, cabling, antennas, air conditioning-related power consumption and site leasing costs are drastically reduced or eliminated. Compared to the price tag (News - Alert) of hundreds of thousands of dollars for a macrocell, the cost per microcell will be in the thousands. There will of course be many more microcells than macrocells (most estimates put the ratio between 5 and 10 to 1). The net effect of this is that for the deployment costs of the microcell architecture to be cost neutral, the cost per microcell needs to be much lower than the cost of the macrocell. This does not mean that the cost of each base station and backhaul link needs to be correspondingly lower than the equivalent macrocell deployment, but that the total microcell solution needs to be similar to the total cost of ownership of a macrocell solution. Given current costs estimates for integrated microcells, that target could be achievable, depending on important variables such as the cost to rent space on non-traditional structures such as lamp poles. This makes microcellular architecture attractive not only for underlay applications to increase the network capacity, but also in greenfield applications where macrocells would otherwise provide sufficient capacity.

The other significant factor to keep in mind is the equipment vendor’s share of total spend increases by a factor of 2 to 3 when moving from the macrocell deployment to a microcell deployment. This is due to pre-integration of all the power and switching systems and elimination of much of the civil and air conditioning-related power requirements. That, coupled with the large potential application space for the microcell architecture due to the attractive operator economics, will have the effect of significantly increasing the potential revenue to the equipment vendors. Even if the microcell architecture is only deployed in metro cores to handle increased network traffic the increased share of wallet will result in a much higher bottom line impact than the equivalent spend on macrocell architectures.

Despite the hype around small cell architectures, there are still a number of technical and operational issues that will need to be resolved before they become a mass market reality. The good news is that the economics seem to be very compelling – both for the network operator and the equipment vendor.