Small Cells: The Next Generation

Posted by Keith Day

Here’s an article by Pete Keevill, co-founder and vice president of engineering at Ubiquisys, exploring the latest developments in the small cell industry and how Ubiquisys and Texas Instruments are working together on the next generation of small cell hotspots. This article was originally published in the latest edition of Texas Instruments’ Insights magazine.

With their sophisticated self-managing capabilities, small cell base stations can be deployed almost anywhere with minimal planning or management, using virtually any IP connection for backhaul. These qualities, together with some powerful market forces, have made small cells a hot topic in the industry, with most mobile operators talking openly about deploying small cells in huge numbers to complement their macro network.

Let’s explore some of the market forces that are driving small cells.

Consumer adoption of smart devices, especially smartphones, is growing to such an extent that they are becoming universal. Despite the use of WiFi in most homes, the consumption of mobile data has relentlessly increased, as mobile app and media usage becomes mainstream behavior. Just how mainstream this behavior is becoming is illustrated by the fact that its growth is immune to serious economic downturn across the world. Previous consumer adoption trends showing similar resilience are AM radio, television and the Internet.

Mobile Data UsageGraph from KPCB Internet Trends

Just as the tide of mobile data usage is rising, the patterns of usage are changing as well. Many app and video sessions are stationary and the usage is location and context specific. The major growth is in indoor public spaces – stores, cafes, transport hubs, malls and public buildings. As a result, mobile operators are facing the challenge of a rapid increase in mobile data consumption that’s here to stay, so they have established several techniques for adding capacity to their networks:

  • Increasing macro layer capacity
    This is the traditional planned approach to data capacity expansion, but by their nature, macro cells are suited to covering outdoor spaces. Also, siting is an issue, especially in metropolitan areas where the maximum capacity is already being reached at existing sites.
  • Increased mobile spectrum
    Regulators around the world are busy re-farming under-used parts of the spectrum (i.e. re-purposing analog TV spectrum to allow them to be used for mobile data). However, spectrum is finite and cannot be brought on stream quickly, and often requires a new generation of handsets.
  • New, more efficient cellular technologies
    So-called 4G technologies like the 3GPP’s Long Term Evolution (LTE) make more efficient use of spectrum, but the capacity gain is small compared to the forecast increase in data usage.

All of these techniques work and all are being employed right now. Together, they can increase mobile network capacity by as much as five times. Unfortunately mobile data consumption is set to rise by 16 times over the next five years, and the cost of multiplying data capacity using these techniques is unsustainable for operators working on today’s consumer pricing margins.

Operators are also turning to WiFi. Most smartphones have WiFi and it already takes most of the load in homes. Operators are working hard at emerging developments of WiFi technology that help overcome the sign-on and security problems that inhibit usage in public spaces. Smart device users tend to be more transient than laptop users, so these concerns are more of an inhibition.

It is into this challenging environment that small cells emerged, and they appear to have the ideal characteristics to solve the data capacity problem long-term:

  • Self-managing, self-organizing
    Small cell radios are powered by sophisticated software systems that sense and adapt to changes in their environment and work cooperatively with their small cell and macro neighbors. This drives down planning and management costs, and enables them to be deployed in a denser, more spectrally efficient formation.
  • Flexible IP backhaul
    Small cells can use any IP connection for backhaul. Indoor public spaces tend to be well-served by DSL or fiber, neatly matching areas of maximum mobile data growth to areas of abundant small cell backhaul.
  • Strong alignment with WiFi hotspots
    Small cells can easily be co-located with existing WiFi hotspots, or the technologies can be combined in a single device, in which case the small cell can integrate WiFi capacity.

So, small cells are the only technology that can provide the huge volumes of mobile data required, and they do so at a sustainable cost. At the same time, the fact is that not all small cells are the same, and most of them are not ready for public space deployment because they are traditional pico or micro cells that lack self-organising capabilities, or they cannot cope with the traffic demands of a real public space environment.

Fortunately, Ubiquisys and Texas Instruments have been working together to develop a solution to this problem.

Ubiquisys has pioneered many of the key technologies required for successful femtocell deployment including “zero-touch” deployment, self-configuration capability with near-continuous “network listen” and “open mode” deployments where the femtocell is used to provide public service with unrestricted user access requiring the handling of unpredictable surges of signalling and traffic. As it became clear that operators were asking femtocell technology to “grow up” to become small cells including both WCDMA and LTE, Ubiquisys was able to draw on this extensive experience to define the key criteria for a successful small cell chipset solution. The headliners are:

  • High Layer 2 processing performance
    3GPP protocols are complex and, to pump a lot of data (around 200Mbps for both HSPA++ and LTE simultaneously) with demanding real-time constraints, takes a lot of compute power just in data packet handling. Smart DMA mechanisms can help here but the maintenance of many packet queues is still a major burden.
  • Layer 3 signalling processing “firewalled” from Layer 2 loads
    Small cells providing open mode service in busy traffic “hotspots” in public spaces or within larger enterprises will have to handle unpredictable signalling surges as users move through the cell while maintaining high traffic throughput. Separating L2 and L3 processing in different processors provides some level of guarantee that signalling is never “starved” of compute power regardless of user traffic load.
  • Backhaul IP processing separated from core “stack” processing
    Small cell backhaul security is maintained through IPSec tunnels. The total processing requirement for IPSec goes far beyond the core encryption/decryption processing which is hardware-accelerated in all femtocell solutions. Throw in any more sophisticated requirements for local offload like Iuh LIPA or packet inspection and the IP processing demands can become a real headache. Dedicated network processing functions separate from the main processor can make this problem manageable.
  • Modem software is a complex, highly skilled task in its own right – and a critical part of the solution

Ubiquisys evaluated a number of leading vendors for a WCDMA/LTE dual-mode small cell solution and selected TI’s TCI6614 SoC as the best in class offering. Based on TI’s KeyStone architecture, the TCI6614 provides specific hardware acceleration for Layer2 including queuing engines and powerful DMA, separate L3 processing and a dedicated network processing engine to handle complex backhaul packet processing. TI’s selection of skilled modem partners for WCDMA and LTE and its longstanding experience in the infrastructure space make them an ideal partner for Ubiquisys’ ambitions in the rapidly emerging small cell segment.

Post to Twitter

This entry was posted in Technology Innovation, Ubiquisys News and tagged , , , , , , , . Bookmark the permalink.