In-building wireless testing can be more complex than outdoor network testing because there are more potential sources of interference, and more variables to coordinate. An indoor small cell environment can include multiple radios and antennas; when these are aligned they can help support one another, but when they do not communicate they can interfere with one another.
Interference with the macro network is another potential hurdle for indoor small cells. According to analyst Joe Madden of Mobile Experts, deployments that are initially independent of the macro network typically come online faster.
“The key is to divorce the in-building system from the macro network for the initial deployment, because the interaction between the two systems can be very complex,” said Madden. He thinks that if indoor deployments can be tested and optimized for different operators after deployment, the outlook for multi-operator small cells may improve. Madden noted that optimization after deployment is a model that has been successful for many DAS deployments, and he predicts that small cell installations will start to follow this model.
“DAS systems are generally deployed and then optimized for each individual operator to suit the unique characteristics of each operator’s macro penetration into the building, essentially adjusting the power levels at individual antenna locations for specific frequency bands,” said Madden. “Eventually, all multi-operator indoor solutions will need to implement this kind of process, with flexibility to adjust power levels for each operator at each frequency band.”
Backhaul is another challenge for in-building wireless testing of small cells. Scott Sumner of Accedian Networks pointed out that inadequate in-building backhaul can sabotage network performance, and said mobile operators may get the blame for equipment that they don’t own or control.
“The biggest problem we have seen with indoor small cells relates to the non-deterministic nature of backhauling over corporate LANs, and then often, the open internet,” said Sumner. “Latency and packet loss can quickly interfere with control-plane messages, so that hand-offs can fail. We have to remember the operator is being blamed by the user if the quality of experience falters, despite their lack of control over the backhaul in many cases.” Sumner added that operators will need to pay special attention to in-building wireless testing of small cells as they roll out voice over LTE.
“Initially, an assessment or audit of the performance of the existing backhaul and RAN is important,” he said. “By evaluating latency, packet loss, availability and throughput the operator will be able to predict whether voice calls will sound good, and whether the sync between sites allows for seamless hand-offs. Most early VoLTE trials and regional deployments failed because the network was not ready, properly configured or insufficient. If the proper class of service (QCI) is not established end-to-end over a bearer, the calls will not only sound bad, they will just stop. VoLTE is much less forgiving that circuit switched voice. It either works great, or not at all.”
For test equipment makers like Accedian, Exfo, Anritsu and Rohde & Schwarz, small cells represent a new opportunity that must be met with new solutions. The equipment used to test the macro network is too expensive and time-consuming for effective use in a small cell environment.
“If an operator chooses to start with the procedures they’re using to deploy and turn up a macro cell, they quickly find out that this isn’t a scalable solution to all these tiny end points out there,” said Sumner. “Really the ROI of the business case of small cells can’t afford to have someone there for a full day doing a turn up test or carrying very expensive test equipment and these delay the rollouts quite a bit.”
Indeed, delays and cost overruns have plagued many small cell deployments to date. But operators and vendors are learning, and solutions are evolving.
“Single-operator small cells have been slow to ramp up over the past few years, partly due to the long time required for planning/installation/provisioning/testing/optimization,” said Madden. “However, some solutions have now emerged with very rapid in-building deployment cycles.”
Efficient test equipment is one major reason the deployment cycles are getting shorter. Vendors are working hard to combine lower-cost equipment with the software needed to optimize multiple small cells.
“You’re trying to find a way to get the least expensive piece of hardware out there that still has the precision to do these complex tests without having power supply requirements, racking requirements and the cost of deployment,” said Sumner. “That’s really how instrumentation has evolved to address these many end points today.”
Follow me on Twitter.
