By Narayan Raman, Yadhav Krishnan, Miguel Hernandez & Furquan Ansari, Bell Labs Computing/Nokia
Software Defined Wide Area Networking (SD-WAN) services are gaining increasing traction in the enterprise communication market because of the confluence of two developing trends – growing cost and complexity of WAN connectivity services, and emergence of virtualization-enabled network programmability. Enterprise IT managers must contend with increasingly complex branch office communication needs being shaped by significant bandwidth growth, emergence of cloud services, proliferation of end points with diverse connectivity requirements and the need to ensure adequate WAN protection against emerging threats, while holding firmly to the expense budget.
Simultaneously, however, network programmability based on policy-based routing, network function virtualization, and analytics-driven zero-touch automation has greatly simplified traditional WAN management. SD-WAN captures these functionalities through a range of alternatives that tailor the WAN solution to specific enterprise needs cost effectively. Enterprises can move low-priority sites from Multi Protocol Label Switching (MPLS) to internet-only connectivity, augment MPLS with internet to optimize local traffic distribution at selected sites or deploy internet breakout at some sites for direct cloud connectivity. SD-WAN also enables flexible branch networking with deployments varying from “thin” end points, which focus primarily on connectivity, to “thick” ones that additionally provide value-add services.
While SD-WAN benefits from the complementarity of these two trends, an MSO needs to evaluate the resulting economics for itself considering its target enterprise market and the required networking and platform expenses. To this end, after briefly describing current enterprise WAN challenges and discussing how SD-WAN addresses them, this paper focuses on how a typical MSO can develop SD-WAN’s revenue and profit potential projections and enhance them through sensitivity analyses of key demand and cost parameters. A distinctive aspect of the proposed model is its utility-based approach that correlates an enterprise’s SD-WAN adoption to its value – measured in terms of desired WAN attributes – relative to its current WAN service. We present an illustrative case study for a generic MSO; the results indicate that SD-WAN can significantly impact the operator’s enterprise revenue while delivering an attractive business case with a payback period of 2.25 years. The proposed modeling approach is extendible to other operators, regions and markets worldwide.