Confronting the Mobile Data Explosion – Where Do We Go from Here?

Filed in Big Data by on June 27, 2013 0 Comments

It’s no secret that the explosive growth of data consumption on mobile networks is presenting a tremendous challenge to mobile service providers worldwide. The broadband services industry – in particular, the mobile data services industry – is working hard to address the complexities of managing capacity within networks to ensure an increasingly high quality of experience for end users.

Indeed, to cope with this massive data expansion, huge investments have been made in the areas of cellular backhaul, spectrum allocation, LTE deployments for greater radio access network efficiency, deployment of Evolved Packet Core for more efficient data transmission in the core, and, of course, external network alternatives like Wi-Fi offload. In this post, I’ll take a look at some of the key challenges and potential solutions involved with trying to manage the data explosion.

First, there are some underlying complexities of measuring and managing packet data flows today. It’s important to “dimensionalize” the network to be able to deliver target broadband speeds during times of peak network demand. Data flows are far more difficult to characterize than traditional voice traffic, and for broadband packet data flows, there is much more uncertainty. Unlike voice telephony, multi-application IP traffic is quite complex, multi-dimensional and dynamic due to the minute-to-minute and even millisecond-to-millisecond changes that can occur in its characteristics.

As a result, network planning and engineering for broadband Internet-type services come with higher degrees of uncertainty because of the following factors:

  • Each application used during an Internet access session, such as video streaming, interactive applications, voice, Web browsing, etc., has different traffic characteristics and bandwidth requirements.
  • End-user devices and applications are evolving continuously at the rate of silicon electronics development, as opposed to voice development.
  • Broadband Internet access supports many different user applications and devices, from streaming high-definition video (unidirectional, very high bandwidth), to short messaging (bidirectional, very low bandwidth).
  • Devices are continuously running many of these diverse applications and services in the background all at the same time, compounding the challenge of monitoring and managing predictable data streams.

Having identified those challenges, let’s examine some of the strategies being used by network operators to address them:

  • Investing in LTE to achieve better bits-per-second-per-hertz (spectrum and radio access network efficiency).
  • Deploying Evolved Packet Core networks that “packetize” all services (including voice) to more efficiently converge and transmit data over larger, more efficient pipes.
  • Interconnecting packet core networks to other packet core networks through IPX network services that are designed specifically for multi-service and multi-applications compatible with an all-IP packet core.
  • Using IPX as a means to deliver high-value data flows between packet cores and external OTT applications for assured delivery and managed flow between core and OTT applications.
  • Deploying data optimization services to provide a better end-user quality of experience while actually reducing packet transmission load on network elements.
  • Deploying real-time intelligence tools within the network to detect and manage service interruptions proactively.
  • Integrating real-time intelligence tools with policy control mechanisms to manage data flows on a per-user basis or even on a time-of-day basis, to cope with peak demand or service logic.

This is just a small sampling of the issues related to ever-increasing data consumption and the necessary tactics for coping with these trends. As with any challenge of this scale, there will not be a single solution. Rather, what will be needed is an entire framework of solutions operating in different places within the network, and solutions capable of traversing network boundaries in an interoperator fashion to ensure end-to-end quality of experience for end users.

Even though much remains to be solved, I am certain the strategies outlined above can provide a foundation for managing the data explosion both today and well into the future. I am curious to get other views. Are we on the right track, or has technology created a data-eating monster destined to outpace network capacity for years to come? I look forward to hearing your thoughts.

Tags: ,


About the Author ()

BJ Neal is Vice President of Network Strategy and works at Syniverse's headquarters in Tampa, Florida.

Leave a Reply