Femtocells, like many new technologies, are maturing and overcoming the technical obstacles which have arisen during trials and testing. Increasing intelligence and self-tuning of these smartboxes ensures minimum disruption of the existing macrocellular network. RF propagation studies have shown the impact and feasibility of high traffic levels in areas of heavy takeup.
We are now seeing a few vendors highlight the importance of appropriate testing capabilities for femtocell development and deployment. These include industry stalwarts, such as Tektronix (and no doubt, Rode and Schwarz and others) as well as smaller upstarts who have identified specific needs.
Tektronix announced their G35 solution for developers in their labs. We don’t really see much distinction between the test equipment available and required for femtocell developers compared to full-size standard basestations. Both need to meet the same specifications and operate using the same interfaces, now that even large cellsites are beginning to use IP connections for the backhaul connections to the main switching centres.
One area that does differ though is the use of relatively uncontrolled, best-effort, broadband service between the femtocell and the operator’s core network. Where operators do have their own broadband service (and some commentators believe that femtocells would mostly be sold with the operators own broadband), then they should be able to engineer priority for voice traffic above that of best-effort data services. Where 3rd party broadband services are resold, then this may not be the case and poor quality service might arise in pockets of overloaded services.
Epitiro have launched a suite of products which provide end-to-end measurement and reporting of the actual performance for these broadband links. By conducting medium term tests over a period of weeks, they can identify if contention on the broadband link causes outages or poor quality at certain times of day. Whilst we might disagree with their suggestion that such testing be conducted for 2-3 weeks before the customer places their order (customers tend to like instant answers when ready to buy), we would agree there is a place for longer term broadband service quality measurement and analysis.
We believe that in order to avoid femtocells getting a bad reputation with early deployments, that voice call quality needs to be high. Customers will expect it to be as good as outdoor service and are unlikely to accept substandard quality. Radio performance within the home should be excellent, which should provide a noticeable improvement. If this is degraded by outages or significant number of dropped voice packets, then this would offset the benefit and cause customer dissatisfaction.
Operators will therefore require the ability to test and measure customer performance both end-to-end (ie from the handset to the core network) and from the femtocell through the broadband link. Differences in these measurements can then identify if poor quality is due to bad RF performance (e.g interference, distance from the femtocell, type of building materials etc) or bad broadband performance (contention, link outages). Appropriate action can then be taken to resolve the diagnosed issues.
Where operators typically have invested in “drive testing” performance measurement equipment for their outdoor network, we would expect similar investments for femtocell performance measurement that provide complementary analysis. This will allow operators to benchmark their service quality and determine what improvements they should demand from their femtocell vendors and/or implement in their own operating procedures.
We can expect to see further developments and announcements in this space.