White Boxes and 5G Mark a Departure from Physical Box Rigidness, but at What Cost?

Network appliances based on open-source software and commoditized hardware are already an integral part of enterprise networks and data centers. Branded as white boxes, they are now emerging as a popular option in the telco domain. A case in point is AT&T, an operator that has gone live with its white-box cell-cite routers, intending to deploy more than 60,000 white boxes in its cell towers over the next few years. The potential utility lies in cost efficiency, network design flexibility, and faster innovation.

White-box utility notwithstanding, technical differences between the strictly regulated and standardized mobile telco network (including 4G Evolved Packet Core (EPC) and 5G Next-Generation (NG) Core) and the open nature of data networks (Internet) will arguably pose many challenges when porting enterprise and web concepts to the telco domain. Further, white boxes have the potential to upend market economics, but the telco industry should first ponder how to fund the Research and Development (R&D) for the required software. Telco vendors have substantial investments in established product lines and their legacy software or hardware, which are capabilities that continue to yield revenue. An imminent diffusion of white boxes, however, may well mean that a telco vendor will be facing a strategic challenge of the first order: what new capabilities do vendors need to compete, and more pressingly, what resources do they need to possess to be able to compete in the future?

At a conceptual level, white-box functionality consists of two parts: the routing functionality and the service intelligence, previously located within the same physical box. The software runs on commoditized hardware, thereby enabling telcos to implement software instances that best support new growth initiatives. A development that abstracts proprietary network components into generic compute nodes, white boxes further pursue a course first introduced by data-driven Long-Term Evolution (LTE) networks where mobile packet core network functions are currently being virtualized and migrated to data centers, or in other words, the private telco cloud. The software in LTE, however, was customized for cellular networks, meaning that the core network Virtual Network Functions (VNFs) remained largely telco-grade, proprietary software based on legacy product lines. Analogously, the rise of white boxes and the advent of 5G will require cloud-native computing, where network components are flexible, scalable, and can follow requirements in near-real-time.

5G’s role to shift telco functionality into the Internet architecture is further emphasized by the 3GPP’s Release 15 in which, unlike previous specifications where interfaces between network functions are represented by box-specific “reference points,” the notion of “service-based interfaces” is introduced to denote interaction between Control Plane (CP) network functions. Clearly, this is an architectural refinement more conducive to cloud-native processes, and potentially another indication of a future in which mobile networking is reduced to a mere plug-in of the Internet. What does this mean for the wider telco community? It currently owns the spectrum, the antennas, and the hardware, so a natural development is to own the software. By its very nature, software favors “services,” which puts telcos in a position to compete with alternative offerings, but on the condition that telcos’ personnel attain the necessary software engineering expertise and know-how.

White boxes first came into existence with Customer Premises Equipment (CPE) devices, an early NFV use case and one that focused on enterprise CPE, specifically for Small-Medium Enterprises (SMEs). This use case is now expanding to include unlicensed, private cellular networks. For a full entrenchment into licensed spectrum, telcos must first asses the technical feasibility of whether Information Technology (IT)-grade white boxes, supported by software, can deliver telco-grade availability and services, underpinned by hardware.

Convergence enables new business models based on customer relationships, but if telcos are not ready to capitalize on this opportunity, it may well be a costly market advancement for them. Telcos should view this development as a harbinger of the new age and as a representative of the kind of bold action needed to capitalize on this incipient industry change that amalgamates proprietary telco system with modular Internet architectures. Pure modularity and proprietary architectures are two ends of a spectrum, and ABI Research expects that most telcos’ architectures in the near term will fall somewhere between these extremes. The cost (and control) of developing the underlying software for future hybrid platforms that bridge telcos and Internet domains remains a challenge. Telcos may consider a “Red Hat” type model where the platform is based on open-source software and support is paid for as an option to capture the growth of tomorrow, but this transition must happen in incremental steps, in order to make the transition as smooth as possible.

If white-box networking takes off, it will certainly alter the competitive landscape for infrastructure vendors. On one hand, it opens up opportunities for niche players that can compete on cheaper price points and convenience. Ambeent, with its software-defined and cloud-based AI powered networking technology is a case in point. On the other hand, some degree of modulatory (or a lack of telco-specific feature differentiation) heightens intermodal competition. Amazon’s white-box foray is an example, albeit one currently aimed at enhancing its positioning in the enterprise hybrid-cloud space where Microsoft dominates. If Amazon turns white boxes into a mass-market opportunity, it may speed up consolidation of the hardware community.

ABI Research expects telcos’ spending in hardware-based networking equipment to continue in the near term, but vendors must plan now to find new growth. There are essentially two ways to achieve this end result. First, venture into new growth areas (increasing scope), as ABI Research has discussed in a previous foresight (Telco Infrastructure Vendors Need to Ride a Wave of Creative Destruction to Address Industry Verticals). Secondly, do more of what is being done (increase scale), but do it in a fashion that reduces cost for telcos. Vendors should infuse their current network equipment offering with technologies that promise operational efficiency. Huawei, Nokia, and Ericsson are incorporating Artificial Intelligence (AI) in their basestation offerings in a bid to automate processes.