Balancing Cost versus Redundancy for Reliable Enterprise 5G Networks

When comparing network requirements between the consumer and the enterprise domains, one aspect has become apparent: above everything else, enterprises (manufacturers, warehouses, shipping port operators, and energy producers) require highly reliable network connectivity throughout their premises. While 5G promises five-nines reliability and network availability (as part of Ultra-Reliable Low-Latency Communication (URLLC) capabilities), enterprise deployments require additional considerations to maximize the degree of reliability and availability, even if certain parts of the network infrastructure break down.

Enterprise deployments will require redundant network infrastructure to guarantee this. While these considerations are far from new, until recently, there has been no clear idea as to the quantity of redundant equipment that is needed to guarantee the desired degree of network availability and reliability for enterprises.

In a recent webinar, German manufacturer of test and measurement equipment Rohde & Schwarz offered an additional model for enterprise deployments, which assumes that from every point in a factory environment, four access points need to be reachable by any connected device. This translates to roughly one access point deployed every 20 meters on a factory floor. While there have been assumptions about network infrastructure necessary for reliable enterprise networks from companies likes Ericsson, Nokia, Huawei, and ZTE, Rohde & Schwarz in this webinar present a viewpoint that starts from the implementing enterprise’s point of view and their requirements toward cellular connectivity. Rohde & Schwarz assumes a similar approach to enterprises in the industrial domain, so it might be closer to reality than any of the previous models.

If reliable enterprise networks require 4X the amount of access points that consumer deployments will require, the Total Cost of Ownership (TCO) for enterprise 5G deployments would increase exorbitantly, which would deter enterprises (specifically small and medium-sized ones) from initial investments. As a result, enterprise 5G deployments will need to change from deployments in the consumer domain in order to mitigate this increased TCO and, therefore, keep the entry barrier for enterprises comparably low. There are two possible scenarios to achieve this.

Enterprises will need to compromise on network reliability. This might be a short-term option to lower initial investment into deployment projects. Taking into account, however, that digitizing mission-critical applications (such as mobile control panels for production machines or remote operations for miners and energy production sites) is a key component of the long-term value proposition for 5G deployments in the industrial domain, compromising on network availability and reliability will not likely be a long-term option.

As a second option, the way in which networks will be deployed might change. Traditionally, the deployment of a communication network on, for example, a factory floor has been a singular event, which will turn into a continuous deployment project that will be gradually enlarged. To keep initial investment as low as possible (which will remain a particular concern in times of tight budgetary controls in response to economic uncertainties around the outbreak of Covid-19), enterprises will start their deployments with a very confined coverage areas.

Combining both considerations, that 5G deployments within the enterprise domain follow a phased approach, which will transform network deployment into a gradual process and will take several years to be completed, the process will follow these very broad phases:

• Phase 1: Deploy a network in a confined area (either a trial project or a well-defined area in a real-life production environment), where no stringent requirements toward availability and reliability are necessary. This would apply to back-office applications (such as asset tracking).
• Phase 2: Densify the existing network infrastructure. Creating redundancy in terms of access points and high enough network availability and reliability can guarantee supporting mission-critical applications. This would allow the cellular network to target front-end applications (e.g., robotics or remote machine control).
• Phase 3: Enlarge coverage area of the 5G network depending on the respective use case requirements. This network coverage enlargement might require a different number of infrastructure components.

Phase 2 and 3 in this model are interchangeable, and which one comes first will depend on the individual priorities of the respective enterprise.

This change in the length of deployment projects underlines ABI Research’s firm conviction that bringing cellular connectivity to enterprises is a marathon, not a sprint.

For infrastructure vendors, this transforms enterprise deployment projects into a gradual process and opens up new opportunities for upselling, as enterprises will continue to enlarge their deployment projects. For network operators and System Integrators (SIs), this means that enterprise deployments will gradually grow in terms of the complexity of network operation.

In addition, the considerations in the section above underpin the need for telco industry actors to understand the technology complexity that deploying a 5G network within the enterprise verticals will entail. However, they will also need to hide the exact same technology complexity from the implementing enterprise. The average small or medium-sized business will have neither the staff nor the financial resources to understand this complexity.

From an economic point of view, different players within the telco industry need to develop business models that reflect the fact that enterprise deployments will gradually become more complex.

These offerings should be designed to keep the necessary amount of initial investment (the key barrier for enterprises to deploy a cellular network) low and increasing with growing complexity. It is important to note that, within the enterprise domain, complexity does not only depend on the number of devices to be connected and the number of access points required. In addition to these infrastructure-focused concerns, complexity in the enterprise domain will also stem from the fact that use cases will become more demanding in terms of requirements for network performance. Most notably, this will be the case, as soon as 5G connectivity expands to digitize mission-critical applications, which depends on extremely high network availability and reliability.