.svg)
While Wi-Fi 7 and private 5G promise lower latency, higher throughput and more robust connectivity compared to traditional wireless technologies, performance heavily depends on the underlying network design choices.
Suboptimal coverage, high interference, cut-offs during roaming and limited capacity are typically a result of insufficient or faulty network design that can hamper the network’s ability to deliver on its potential.
This week’s Zero-Touch newsletter guest author is Jussi Kiviniemi, founder and CEO of Hamina Wireless. Jussi shares his perspective on enterprise wireless network challenges and key design considerations, including coverage, capacity and roaming. He also highlights key aspects of enterprise wireless network design and important considerations.
While public 5G networks cover most outdoor areas on the planet, indoor wireless is a different story. Energy-saving building materials like thermal glass prevent heat and cold from traveling in and out of the building but simultaneously block or degrade wireless signals that attempt to penetrate them.
Wi-Fi has become the de facto technology complementing public cellular for high-speed data transfer indoors, supporting 70-80% of indoor wireless data transfer today.
However, in specialized use cases such as warehousing, factories, airports and shipping yards, private 4G and private 5G networks have also become an increasingly popular part of the mix.
High-quality Wi-Fi and private cellular faces expectations for flawless coverage and demands for higher network capacity that skyrocketed as bandwidth-intensive video conferencing apps like Zoom and Microsoft Teams were popularized.
These developments demand higher density networks but without proper design, a dense setup of access points and base stations can cause self-interference and end up performing worse than a sparse one.
Imagine being tasked with standing up a flawless indoor network in every building in the city. That’s literally the remit for Wi-Fi today. Every business building in the world has Wi-Fi, and now an increasing amount are also adding private 4G and private 5G.
Every day, thousands of networks are designed and deployed—often without careful consideration of use case requirements. The result? Dropped calls, choppy video and disrupted business operations.
The massive scale of enterprise indoor wireless networking demands networks be designed and deployed quickly at minimal costs.
This is achievable, provided the network design tools that reduce labor hours and costs fit within budget.
Public cellular networks are designed by dedicated, experienced telecommunications and RF experts. These hardcore engineers design networks day in and day out.
On the other hand, wireless enterprise networks are often designed by generalist network engineers with switching, routing, network access control and even run-of-the-mill IT support expertise.
This means many of the network engineers designing and deploying Wi-Fi (and to some extent, private cellular) won't have a dozen years of experience in wireless or a Ph.D. in radio frequency mathematics. There are fairly few hardcore-Wi-Fi engineer experts in the world. For example, the number of Certified Wireless Networking Experts (CWNEs) has only recently surpassed 500—far too few to design and support every enterprise worldwide. While every enterprise is different, most businesses are at least somewhat dependent on solid wireless networking connections.
Private cellular networks have dedicated spectrum but often with limited allocation. This means high reliability but lower network capacity. For example, the lightly licensed CBRS band in the US provides a total of 150 MHz, some of which is allocated for each user per region.
Wi-Fi on the other hand provides 1,200 MHz of spectrum for the new 6 GHz band alone in the U.S. and about 500 MHz in the EU, with hopefully more in the works. This is in addition to the hundreds of megahertz available on the traditional 2.4 GHz and 5 GHz Wi-Fi bands combined. While high capacity, because Wi-Fi is license-free, anyone can set up a router next door that may interfere with your enterprise network. The newly allocated 6 GHz spectrum for Wi-Fi not only provides massive amounts of additional capacity but also lower latency due to new, clean and wide 6 GHz spectrum, and smart protection mechanisms for avoiding incumbents on the spectrum, such as point-to-point links outdoors.
The 6 GHz Wi-Fi space means a better life for new 6 GHz capable Wi-Fi handsets and legacy Wi-Fi devices still continuing to utilize the traditional 2.4 and 5 GHz Wi-Fi bands. As new handsets like laptops and phones move over from legacy 2.4 and 5 GHz bands to 6 GHz, they free up more of the 5 GHz airtime for the devices that stay there.
With the challenges defined, now let’s explore some of the fundamentals of wireless enterprise network design.
Accurate coverage planning begins with obtaining high-quality floor plans or outdoor environment maps to accurately model the radio propagation. In an indoor scenario, the design must account for obstacles such as walls, warehouse shelves and preferably even signal bleed between floors. The building structures attenuate signals differently based on composition, thickness and layout. In outdoor environments, terrain contours, foliage and buildings create additional complexities in coverage modeling. To aid the network designer, modern tools, plain bitmap images, PDF files and CAD files can all be turned into 3D RF designs in a matter of minutes. Many modern outdoor-targeted solutions also import clutter and terrain data automatically.
Designers should evaluate primary and secondary coverage areas: primary coverage ensures each area receives the target signal levels, while secondary coverage acts as a buffer for load balancing and handovers. Simulation of signal propagation, penetration losses and reflection patterns are essential for designing Wi-Fi and private cellular networks to maintain robust connections across the entire coverage area.
Enterprise Wi-Fi uses access points wired to a switch, unlike home setups that often use wireless mesh networks. In contrast, private 5G networks use gNodeBs or eNodeBs to provide coverage, depending on whether they operate in 5G standalone or non-standalone mode.
The choice of indoor or outdoor equipment, and specifically whether to use internal or external antennas, is driven by factors such as the environment, coverage demands and aesthetic constraints. High-gain directional antennas may be necessary for large or complex layouts to ensure adequate signal levels, while omnidirectional antennas can be used in more open floor plans – or traditional carpeted space environments. Selecting the optimal antenna type, tilt and placement affects both coverage and interference control. Antennas that are poorly aligned or in suboptimal locations can create coverage holes or strong interference zones, undermining network quality.
Beyond signal coverage, capacity planning is a key factor in RF design for Wi-Fi and private 4G and private 5G. Determining how many carriers or channels can be deployed—and at what bandwidth—directly affects how many simultaneous users or devices can be served without performance degradation. For capacity planning, it’s crucial to know how many client devices will be connecting to the network and where. Different areas may have different client densities.
Channel bandwidth allocation and the use of carrier aggregation have direct implications on throughput. As Wi-Fi and private cellular networks must handle mission-critical or high-density IoT device deployments, an in-depth analysis of user density, device distribution, and balancing a reasonable number of client devices per radio is essential. Tools that can factor in interference from neighboring cells help ensure the network can scale to meet peak demands without compromising reliability.
Roaming between cells should also be considered. Designers must plan handover regions to maintain minimal dropouts and avoid abrupt signal loss when devices transition from one cell to another. Overlooking roaming zones can result in frequent disconnections and degraded user experiences. Advanced digital twins can be used to mimic the handset behavior and predict the roaming and handover regions, and overall wireless experience, around the network.
Wherever private 5G is deployed, there’s a good chance Wi-Fi is also needed or already in place.
It’s essential to map out the responsibilities between 5G and Wi-Fi, for example:
AI (or specifically, machine learning) can identify widespread as well as needle-in-a-haystack problems in network operations and also help in network design.
An obvious example is image analysis using machine learning (ML), which is good at picking up objects from floor plans, such as wall structures, doors and windows. While previously engineers have been forced to draw walls and other structures manually (or obtain layered drawings such as CAD files), machine learning can automatically trace the main walls based on even a grainy JPEG format floor plan.
Automated analysis of floor plan structures may sound like a small thing for machine learning, but it may shave off more than 50% of network design time.
In cellular networks, Open RAN has opened the playing field between vendors. Enterprise networks are even more advanced in API maturity and connectivity with other systems. An increasing share of enterprise wireless networks are cloud-operated, and all cloud-operated enterprise networks support a wide range of APIs. The APIs allow network admins to perform automated mass-configurations and deployments of networks using scripting, or through API-assisted third-party network design tools. APIs can also enable third-party tools to create a 3D digital twin of the enterprise network, allowing users to drill down to specific radios or client devices for troubleshooting, even capturing wireless packets and running active client device tests.
It is increasingly common to use APIs and AI together in enterprise networking, blurring the barrier of end user scripting, networking-vendor-provided user interfaces and third-party tools. A network engineer might run a script to deploy new networks without ever touching the network console or dashboard, then use the dashboard to observe AI-based alerts, and drill down on a specific location using a third-party tool’s 3D visuals.
Key factors in designing a great enterprise wireless network that meets customer requirements include acquiring high-quality floor plans, planning for primary and secondary coverage, considering advanced factors such as roaming and signal reflections, and ensuring sufficient capacity, based on client device density and locations.
AI (more accurately, ML) and APIs are not just buzzwords in enterprise networking but already are saving tens of thousands of hours of network engineer time every month.
The most important consideration when designing and building a great wireless network? Thoroughly understand the customer requirements and use cases. Talk to the network users and business stakeholders. This crucial step cannot be automated, no matter how good the AI or APIs.
