According to Maravedis Research, the mobile virtual network operator (MVNO) landscape will soon be dominated by new organizations as cable providers or multiple-system operators (MSOs), over-the-tops (OTTs), device manufacturers and retailers offer their own MVNO, with the most significant opportunity belonging to the OTTs. By 2018, these OTTs could represent over 20% of the MVNO market.
The MVNO Market
Types of MVNO by percentage of MVNO subscriber base.
(Source: Maravedis Research)
In light of this, it is probably no surprise that the rumors are swirling in the last several weeks about Google potentially launching their own MVNO
, in partnership with Sprint and/or T-Mobile.
Rather than argue the pros and cons of Google’s emergence as a mobile provider, I'm going to focus on the MVNO industry and continue my technical rationalization of software-defined networking (SDN) and network functions virtualization (NFV) for mobile networks, this time from a MVNO perspective. (See SDN & NFV in Mobile Networks: The SGi-LAN and SDN & NFV in Mobile Networks.)
MVNOs are not a new phenomenon. They date back to as early as 1997 (for example: Telstra and Bellsouth in New Zealand). Over the last several years, the MVNO landscape has been reinvigorated and reinvigorated in the last several years with the emergence of WiFi as a viable data (and cost!) offload.The earliest MVNOs were built around the assumption that cellular networks were the primary access network, and secondary access networks were WiFi.
Today, we see many MVNOs adopting a WiFi-first model in which the WiFi network is the predominant access network and cellular is the secondary access network. This evolution of both technology and consumer usage trends allows us to question the validity of many of the earliest MVNO business and technology model assumptions.
In general, the MVNO model was built on four core assumptions:
- High-segmented markets for reselling mobile network operator (MNO) services are attractive: OTTs and MSO MVNOs are able to challenge this model by targeting their users and subscribers with differentiated service offerings and pricing models by combining their existing services with mobile services.
- Vertical integration of access and services is desirable to subscribers: Whether we look at video or messaging or voice, we are witnessing an increasing shift of mindshare and wallet-share towards access-independent services. OTTs and MSOs have considerable content and service market share, making their emergence as MVNOs an interesting dynamic for the consumer.
- User-network relationships are long-lived: Many of the important metrics in the mobile industry, such as subscriber acquisition cost, subscriber growth, subscriber lifetime value, subscriber retention rate, subscriber retention costs, etc., are dependent on the assumption that subscribers and network providers maintain multi-year relationships. Part of the stickiness in this model was device subsidization. However, with increasing enterprise BYOD, more device financing and more consumer channels for outright device purchase, the subsidization model is becoming less important.
- User-network relationships are persistent: In addition to the multi-year relationship, there was also an implicit assumption that a mobile device will connect to the same network for long periods of time. In other words, the mobile industry relies on a model that includes a primary access network and secondary access networks leveraged in series. With the proliferation of both public and private WiFi networks and a bit of technology innovation, it is becoming plausible to deliver multi-homed devices that take advantage of multiple available networks, both cellular and WiFi, simultaneously.
As early as 2003, academics have coined the term "Always Best Connected" (ABC) networking and applied the model to MVNOs. This term describes networking based on the premise that the user, or service provider, not the access network operator, decides the best network for the content and delivery of services (M.P.de Leon, A.Adhikari "A User Centric Always Best Connected Service Business Model for MVNO," ICIN2010 proceedings).
Apple even alluded to a similar model in a 2006 US Patent and Trademark Office filing entitled "Dynamic Carrier Selection" (Apple Inc. "United States Patent Application: 0110130140." United States Patent and Trademark Office. N.p., n.d. Web.).
The functional enablers of this model are almost all there -- mobile embedded SIM cards, traffic offload using coordinated selective IP traffic offload, access network selection algorithms and multipath capabilities are all at various stages of standardization, development and inclusion into mobile devices. So, technologically, we have come a long way from academia, but the operational aspects of this model, especially the introduction of significant variance, still remain.
SDN, NFV and MVNOs
Enter SDN and NFV, two technologies and operational models perfectly suited for MVNOs, where efficient use of underlying resources in environments with high degrees of temporal and spatial diversity is of the utmost importance. In short, SDN and NFV seek to remedy the impact of "variable" in both MVNO costs and operations.
MVNOs can leverage SDN as the foundation for separating the rules and topologies associated with mobile network functions, such as the LTE EPC and Gi LAN service elements, from the forwarding functions that connect them even over wide area networks (WANs).
MVNOs can also leverage NFV to eliminate the physical topology boundaries of individual network functions, creating new operational paradigms that allow automation to play a bigger role in capacity and scaling management. Synchronizing forwarding policies with IP flow policies and network selection policies provides MVNOs with the ability to dynamically manage devices, subscribers, services and resources consistently and continuously across multiple access networks.
By leveraging SDN and NFV in the data center and SDN in the WAN, MVNOs can deploy network functions in their cloud, consisting of multiple physical data centers, and intelligently steer traffic to the most optimized location based on any operational or contextual parameter, including time of day, resource availability, device point of attachment, subscriber profile, charging rules, network outages and more.
The same model described for MVNOs applies to machine-to-machine (M2M) providers, as well. Architecturally, a number of M2M providers operate as MVNOs, albeit with different business relationship with the underlying MNO and different economics. Nonetheless, applying the operational principles of SDN and NFV will allow these organizations to shift an increasing amount of intelligence and decision logic out of the infrastructure itself and into customized, custom-built, or homegrown service delivery platforms that interact with the network and network functions.
— Kevin Shatzkamer, CTO, Mobile Networking, Brocade, special to The New IP