The insatiable demand for high speed LTE mobile networks presents incredible opportunity to operators, but it also...
introduces significant challenges. Beyond the need for scale and bandwidth, the move to LTE poses a slew of operational and management issues that SDN and network virtualization may be able to resolve.
The swift expansion of LTE networks is fueled by enormous growth in mobile data traffic and especially video. Cisco's Visual Networking Index (VNI) Global Mobile Data Forecast estimates that by 2017, video will account for two-thirds of all mobile data traffic.
Keeping up with this level of video expansion will require intensely scaled bandwidth, but that's only part of the issue. Video places particularly intensive demands on LTE networks with respect to content caching, transcoding and overall speed. What's more, mobile networks are not monolithic -- they run 3G and 4G traffic, as well as a slew of applications simultaneously, which adds to the management complexity. At a time when revenues and margins from traditional voice services continue to plummet, operators have to be particularly mindful of capital expenses and ongoing maintenance and support cost efficiencies.
To better manage these complex environments, the industry is exploring whether SDN can play a role in LTE networks. Vendors and operators alike have already considered how SDN can be used in WiFi networks, but earlier this year, the Open Networking Foundation (ONF) formed the ONF Wireless and Mobile Working Group and expanded that research to include LTE.
An SDN LTE network could reduce complexity
The ONF Wireless and Mobile Working Group will examine ways to apply virtualization to mobile networking, beginning with three projects: mobile packet core, wireless backhaul, and integrated fixed wireless operations.
OpenFlow, which applies an SDN architecture with a flow-based model and a control scheme that centralizes intelligence and promotes a high degree of network programmability, is a good fit for dynamic and growing mobile networks.
For operators, having a mechanism that allows them to easily configure networks in a centralized manner sounds like Nirvana. While there is still much work to be done, the appeal of a more automated and elegant approach to managing fast evolving LTE networks makes SDN appealing.
For example, SDN architecture might be a good fit to address the increasingly vexing issue of inter-cell interference in an LTE environment. The tremendous increase in mobile traffic has resulted in a big jump in the number of cells in Radio Access Networks (RANs) users rely on to get an entrée to a 3G or 4G mobile network. Many cells are in close proximity of other cells, leading to interference that can quickly degrade service quality.
While there are existing inter-cell interference coordination options to help reduce signal-to-interference plus noise ratios today, they tend to be distributed in nature. The existing techniques add to processing overhead and management complexity. They can also drive up power and network utilization, as well as operating costs.
In theory, SDN can help reduce the complexity associated with interference management by centralizing control of all radio resource allocation across multiple base stations. Consolidating the intelligence and control supports greater resource efficiencies, allowing decisions to be based on dynamic power requirements and the profile of each base station.
SDN will change how mobile traffic is managed
SDN also could be an excellent option in helping operators address the increasing challenges associated with mobile traffic management and the complexities brought on by video.
SDN's inherent emphasis on optimal path routing and traffic steering could facilitate more efficient traffic routing across LTE networks, enabling them to scale better to meet escalating capacity needs while supporting service quality.
For example, using SDN, operators could make dynamic changes to the network based on factors such as individual flow rates or aggregate flows based on specific ports, flow duration, number of users per base station, or by available bandwidth, application or IP address. The factors that trigger a change can be set and reset based on new conditions.
What is clear from just a couple of potential use cases is that SDN could eventually play a compelling role in the LTE networks of the future. Given the massive growth and the increasing dependence on mobile data networks, SDN's capabilities could provide crucial management and coordination capabilities that not only address difficult challenges, but support a more optimal environment for operators and subscribers alike.
About the author:
Amy Larsen DeCarlo has worked in the IT industry for over 17 years and is a principal analyst at Current Analysis for its security and data center services. Amy assesses the managed IT services sector, with an emphasis on security and data center solutions delivered through the cloud, including on-demand application, unified communications and collaboration, and managed storage offerings.
How NFV plays a role in Voice over LTE
The move to 4G could make operators SDN believers
The Open Networking Foundation tackles SDN mobile and wireless
Amy Larsen DeCarlo asks:
Is network management the killer SDN app?
0 ResponsesJoin the Discussion