Insight: Investing in HetNets

Posted by Mae Kowalke on Sunday, April 19, 2015 with No comments

Wireless operators investing in their heterogeneous networks (HetNets) have one overarching goal in mind: future-proofing those investments by choosing solutions that can scale and adapt to new technology. Small cells and distributed antenna systems are an important part of HetNet evolution. 

Martha DeGrasse, Wireless Infrastructure Editor at RCR Wireless News and Scott Sumner, VP of Solutions Marketing at Accedian Networks, explored the roll of small cells and other aspects of HetNet investments during a recent video interview

Some of the questions explored during that conversation are summarized here. 

Why does it typically take 18 or more months from small cell solution sign-on to actual deployment?

Some of the contributors to delays in small cell deployments include:
  • Site acquisitions and tower permits
  • Backhaul setup, especially for distributed antennas
  • Procedures used for deployments
That last one is worth exploring a bit further, because small cell deployments require a different set of procedures than macro cells. ROI featured in the business case for small cells gets cancelled out if a full day of turn-up testing is performed at each site. Often, deployments are delayed as this reality becomes clear and adjustments must be made to procedures and the tools used. 

Where is interference the biggest problem for small cell signal: outdoors or indoors?

Last-mile backhaul, which may include indoor and outdoor resources, is the crux of many interference problems, experienced by users as degradation of service quality (e.g. slow data speeds, poor call quality or dropped calls). 

The good news is that many small cell gateways now are able to detect someone coming into a building, for example, and keep them on the local network to prevent interference with macro cells.

Carriers often don't have control over backhaul, and certainly not the last 100 meters or 100 yards to the desk or small cell antennas that are part of the corporate LAN. Operators are in the unenviable position of being responsible without having control. (This is especially problematic when backhaul involves the open internet.) 

How can operators monitor and assure services that rely on off-net resources?

When off-net resources are involved in backhaul, operators need a way to quickly detect problems and determine their origin. Large operators typically wholesale the last mile, an arrangement that includes an SLA with the backhaul provider and monitoring tools to cover the operator's end-to-end responsibility for the service. 

For end-to-end monitoring and assurance, operators set up a small piece of instrumentation--either a smart SFP or NID--at the customer site. This is used as a tunnel to test connections for compliance with the customer SLA, and locate off-net issues to hold the wholesaler responsible for its SLA. 

Of course, there's no SLA for the open internet. But, with proper testing and assurance tools, at least the operator can identify if that's where a problem occurs, and let the customer know what's happening.

In terms of last-mile service degradation, Wi-Fi connectivity issues are a classic problem. To figure out what's going on, the operator must be able to test the air interface from the end-user perspective, as well as monitor the small cell backhaul connection. Low-cost devices (using NFV) are now available to monitor both from the same location. 

How has network functions virtualization (NFV) changed network monitoring and assurance solutions?

The NFV-based modules now used for end-to-end performance monitoring and assurance originated as smart SFPs with embedded FPGA instrumentation, with a small enough form factor to plug into a small cell gateway router. Over time, these endpoints have evolved with a variety of media conversion options to address factors like the prevalence of copper connections and CAT-5 cabling in enterprise environments. 

Network function virtualization (NFV) can replace the CPU functions of these devices, allowing control to be centralized in a hosted data center. The goal now is to find the minimum, least expensive piece of hardware with the precision to do these tests, without having power supply or racking requirements and the associated deployment costs.

Accedian's SkyLIGHT VCX Controller uses that arrangement for small cells monitoring and other, related applications. Although NFV can be complicated, VCX is packaged in a way that allows installation as a physical or virtual appliance. From the control interface, virtualized modules appear the same operationally as traditional, stand-alone units. 

VCX modules can be self-installed by the customer, and they cost up to 90% less than traditional hardware. 

Once operators move their measurement and monitoring platforms to a VNF-type basis, the development curve is going to be incredibly fast. For example, it took Accedian about ten years to develop products for hardware standalone appliances, and about ten months to reproduce that in an NFV-based environment.

It will be exciting to see what people do with NFV. This is definitely an opportunity for QoE to keep up with the growth in traffic demand.

For more insight into HetNets and mobile service assurance--including what operators are doing to prepare for VoLTE, and why MOS matters--watch the full video of Martha DeGrasse interviewing Scott Sumner.