Feature Article

December 08, 2008

Bigger Isn't Always Better

Whenever new technologies are introduced there are a lot of conflicting statements made about figures of merit for performance, economics, user experience, etc. Usually all of the statements have some degree of truth under certain conditions. The problem is how to sort the wheat from the chaff. The discussion surrounding base station size (Macro, Micro, Pico or Femto) is certainly subject to this same confusing barrage of information and misinformation.
The first order constraint in a wireless network design is reach — how far can you go before the signal degrades to the point that you can no longer deliver the service offering. This is made worse with 4G services and the higher data rates typically require stronger signal strength, resulting in reduced reach. The Radio Access Network (RAN) operating frequency determines how the signal is affected by obstructions such as trees and buildings. The lower frequencies like 700 MHz are less affected by these obstructions and the higher frequencies like 2.3 GHz or 3.5 GHz are affected more strongly. The reach defines the maximum spacing possible for the base stations. Normally the fewer the base stations, the lower the network cost — or so one would assume. However as usual there are other factors to consider.
The second constraint is the amount of spectrum available to the operator. This, coupled with the bandwidth of the service being offered to the consumer determines the number of customers that can be served by a single base station. A bit of math — at 64 QAM (the maximum modulation rate for WiMAX or LTE networks) you get ~ 6 bits per Hz, so in a 10 MHz channel it is possible to deliver ~60 Mbps raw capacity. Subtract some for overhead, signaling and forward error correction and you get ~45 Mbps. Assuming a typical download/upload ratio you would get 30 Mbps down and 15 Mbps up. If you wanted to provide 1 Mbps service to your customer each base station would be able to support a maximum of 30 simultaneous users. Clearly this does not make for a very attractive business case given the rates people are willing to pay for mobile broadband services. In any urban or suburban region this teledensity (the number of subscribers supported per base station) determines the base station spacing — not the reach.
The principal focus of base station design, then, is all about how to maximize the use and re-use of the RAN spectrum. Techniques such as MIMO (multiple input multiple output) and sectorization (using directional antennas operating a different frequencies on the same tower) are used in the macro base station to spatially localize the signal and allow the same slice of spectrum to be re-used more frequently. There are limits to the gains that can be made here and these techniques drive up the cost of the macro base station. Micro or Pico base stations refer to very small base stations that are mounted on lamp poles or the sides of buildings instead of on traditional cell towers. The objective is to allow the operator to reduce the base station size until the teledensity matches the subscriber density by deploying 10 to 100 times more base stations than the macro designs. The challenge is that in order to keep the cost model intact both the base station cost and the backhaul cost have to drop by a similar factor. This drives the need for using the same equipment and spectrum for both access and backhaul and reduces the spectrum re-use gain of this approach. This, coupled with the operational challenges of deploying and maintain such a large number of devices have to date resulted in very limited deployments of this technology. New devices with higher levels of integration and resulting lower cost points are now becoming available which may result in the equation tipping in favor of micro/pico base stations.
Femto cells are trying to solve a very different problem. Here the focus is on in-building coverage, not spectral re-use per se. A phrase used to describe this segment is mobile broadband access at home (MAH). Most of the spectrum, which is being made available for 4G services is in the 2 to 3 GHz range, which has relatively poor ability to penetrate through walls. This coupled with the high signal strength required to deliver the high data rates means that it is very difficult (and expensive) to provide mobile 4G services inside of buildings. The femtocell proposal is to essentially add a base station to the cable or DSL router to provide the indoor coverage while the macrocell provides the mobile coverage. The added advantage of this is that the femtocell backhaul is provided by the existing wireline connection and does not further burden the mobile network costs. Finally, the femtocell is seen as an enabler for fixed mobile convergence — allowing the user to get all of their voice, video and data services from a single provider with a corresponding increase in service levels or decrease in cost. The devil is always in the details and challenges of per unit cost (and the resulting need for subsidies), spectrum co-ordination between neighboring buildings, tariffs from the mobile operator to the wireline operator (if they are not the same company) for the femtocell backhaul, device management, hand-off between femto and macrocells… These issues are being worked out and many operators are considering femtocells to be integral to their LTE strategy.
So why am I discussing base station issues in a column entitled “The Middle Mile?” The answer is simply because the base station design dramatically affects the backhaul demand. The critical factor that determines success or failure of the micro/pico cell model is the cost of the backhaul. Traditional partitioning of the backhaul and base station as separate platforms can simply not be supported by the physical or economic constraints of this model. This may be seen as a threat to traditional backhaul technologies, however the Micro/Pico cell model also delivers 10 to 100 times more bandwidth to the subscriber, resulting in a corresponding increase in backhaul demand. Similarly, the femto cell model offloads the traffic while the user is at home, reducing the backhaul burden on the macrocell. The total network traffic, however increases by a significantly larger factor than this offload due to increased adoption of 4G services. In both scenarios, although the backhaul pie may be split between more solution sets, the total pie is dramatically increased, resulting in good news for all backhaul vendors. The need to continue to adapt one’s product lines to track the changing requirements has never been more apparent. Those who cannot bring themselves to cannibalize their existing products with new offerings or who cling to traditional, TDM based technologies will be left behind in this rapidly evolving new world.

Dr. Alan Solheim, Vice President of Product Management at DragonWave, is author of MobilityTechzone�s The Middle Mile column. To read more of Alan�s articles, please visit his columnist page.

Edited by Greg Galitzine

FOLLOW MobilityTechzone

Subscribe to MobilityTechzone eNews

MobilityTechzone eNews delivers the latest news impacting technology in the Wireless industry each week. Sign up to receive FREE breaking news today!
FREE eNewsletter