Wi-Fi 6 vs. private 5G will drive competitive positioning among vendors
June 24 2020
by Mike Fratto
A battle between Wi-Fi and 5G is brewing, just not out in the open. Most experts agree that 5G and Wi-Fi will be complementary to each other, an assertion we made in February. For example, Wi-Fi will remain a mainly indoor wireless technology, while 5G is better suited outdoors but can also be used inside. However, what's happening at the product and operator levels is anything but complementary. IoT is growing, and Wi-Fi vendors want to be included in those deployments. According to our 2019 IoT Market Monitor, IoT, including consumer and industrial IoT, will have an average five-year CAGR of 31%, growing from $172bn in 2019 to $658bn in 2024 (this forecast has not been adjusted to reflect the impact of COVID-19). Meanwhile, managed service providers want to expand the use of Wi-Fi to retain consumers and businesses. There is an opportunity for 5G in carpeted spaces. Some operators are looking to drive 5G into carpeted spaces beyond merely extending their 5G networks into buildings.
The 451 Take
Wi-Fi and private 5G are complementary wireless technologies, and you would have a hard time finding a vendor claiming differently. However, Wi-Fi vendors will be attacking private 5G vendors that encroach on the enterprise, claiming enhancements to Wi-Fi 6 will make it a suitable replacement for private 5G, including supporting deterministic connectivity, higher performance, better conformance to customer-defined service levels and higher device density. There are a lot of assumptions packed into those claims that need to be validated in live deployments. Even in the best case, Wi-Fi 6 won't be able to guarantee the kinds of performance and density levels private 5G can achieve. Private 5G has a place in some enterprise IT environments, and enterprises ought not be scared off by the new technology.
Use cases are driving wireless technology adoption. Currently, most of the emphasis of wireless IoT connectivity has been on commercial and industrial environments, particularly in the financial, government, healthcare, manufacturing and retail sectors. There are several IoT protocols in use over wired and wireless networks, but most of them are legacy protocols designed for specific purposes and often tied to specific products. Enterprises and integrators are looking to unshackle themselves from custom technologies using standards-based and commonly used technologies, such as Wi-Fi, 5G and 4G, Ethernet, and BLE. According to our Voice of the Enterprise: Internet of Things, Workloads & Key Projects 2020 survey, over the next two years respondents expect their use of 5G for IoT to more than double, while Wi-Fi and 4G/LTE use declines (see figure below).
Which of the following network connectivity technologies does your organization use for IoT connections?
451 Research, Internet of Things, Workloads & Key Projects 2020
The data illustrates Wi-Fi vendors' concerns about being shut out of an emerging and rapidly growing IoT market, and explains the hype surrounding operators offering 5G services. Operators have been looking for ways to generate more business within enterprises outside of WAN and managed services. 5G connectivity for IoT represents such an opportunity. Likewise, equipment vendors such as Airspan, Celona, CommScope, Federated Wireless, Redline Communications, and Nokia are offering private 5G network hardware, software and services. Vendors that are heavily invested in Wi-Fi, like Aruba and Cisco, want to blunt the encroachment of 5G where they can and, more importantly, capture a larger share of the growing IoT connectivity market opportunity.
Strategic vision and business drivers
With these new capabilities in Wi-Fi 6, Wi-Fi vendors will claim that Wi-Fi 6 will be as effective as 5G for indoor use and for IoT at delivering deterministic wireless networking at scale. Wi-Fi vendors will point out that the oft-discussed ultra-reliable low-latency goals of 5G, like less than one millisecond of latency and six nines of reliability – about .6 seconds of downtime per week – are only important in a small subset of specialized use cases, like over-the-air robotics control and autonomous vehicles. Today, time-sensitive applications like real-time voice and video usually perform well enough over Wi-Fi, even in moderately congested locations, which lends credence to the claim. However, without guaranteed clean airspace, time-sensitive applications may suffer degradation on Wi-Fi, which is risky for mission-critical applications like VoIP. The problem with Wi-Fi is that the 2.4Ghz and 5Ghz bands are unlicensed and unmanaged. Anyone can use them, which leads to co-channel interference. Wi-Fi access points (APs) support a variety of channel widths, and users will often select wide channels, thinking they will get better performance, but are stomping all over the band and making performance worse for everyone, including themselves. Wi-Fi 6 doesn't change that situation at all.
Private 5G, on the other hand, uses managed spectrum and offers significant benefits over Wi-Fi. The Citizens Broadband Radio Service (CBRS) in the US has three license tiers, and access to the spectrum is managed though an FCC-approved Spectrum Access System (SAS). Different countries are using different allocation and license schemes. The free General Authorized Access tier is guaranteed a minimum of eight channels in the band. GAA can use more channels if they are not being used by the higher-level tiers. Priority Access Licenses are paid tiers that guarantee a licensee access to that spectrum for a county. Seven channels can be reserved for PAL licensees. The Incumbent tier takes priority over all other tiers, but the impact should be limited to a few regions along the US coastline. Spectrum management brings some measure of control and guaranteed performance to CBRS-based wireless, and is where SAS operators can innovate their services and build competitive differentiation.
Wi-Fi vendors will also claim that their Wi-Fi 6 APs can support a higher density of devices than with previous Wi-Fi versions. This claim is related to the spectrum management and multi-user support, which allows more wireless devices to use the channel simultaneously. Device-density support is important not only for scenarios where there are many devices in a confined space – for example, healthcare with connected medical equipment, and educational settings with staff and student wireless laptops, phones and tablets using the network. Private 5G radio systems can support thousands of devices per radio, but the requirement for a pre-installed SIM card means it is better suited for scenarios where IT has control of the devices (like in healthcare) and Wi-Fi is inherently more accessible for staff, customers, students and guests.
Finally, Wi-Fi vendors will claim that Wi-Fi is less expensive from both a capital and operational perspective than private or public 5G because Wi-Fi 6 APs and the associated hardware and software will be lower-cost compared with new 5G radio stations and the evolved packet core required to run them. Additionally, Wi-Fi is already familiar with IT, so there is no reskilling required, just an update on Wi-Fi 6 capabilities. These are valid claims, and wireless vendors are already publishing educational materials about Wi-Fi 6 – the technology features prominently in their product literature and events. Cellular technologies are largely unknown in the enterprise, which means education and training are required for IT to deploy and manage them. Vendors offering private 5G products, such as Celona, CommScope, Federated Wireless and Nokia, offer extensive training and professional services to enterprise IT and integrators. These vendors are also developing products for the enterprise that will be simpler to deploy and operate than the products available for telecom operators. There will be additional costs for the enterprise to self-manage private 5G networking products, but they may not be as high as some claim.
Wi-Fi 6 brings a number of enhancements to spectrum management, which will increase both the number of Wi-Fi 6 clients that can be supported and the over-the-air speed when devices are connected to a Wi-Fi 6 AP. It offers no new improvements for Wi-Fi 5 and earlier clients connected to a Wi-Fi 6 AP.
Wi-Fi 6 is the first iteration of the wireless protocols from the IEEEE 802.11 working group to tackle active radio spectrum management on the client using orthogonal frequency division multiple access (OFDMA), which allows the band to be allocated into smaller chunks called resource units and 'multi-user multiple input and multiple output' (MU-MIMO), which allows simultaneous use of the band between multiple clients.
In Wi-Fi 5 and earlier, clients would consume the entire band while receiving or transmitting, which limited the use of the medium to one client at a time. Wi-Fi 5 did bring the ability for APs to transmit simultaneously to multiple clients, but it wasn't particularly effective. Wi-Fi 6 radio management uses the AP to collect information from clients, like how much data needs to be sent, and the AP can then allocate – dynamically and in real time – a section of the band to multiple clients to transmit. This means multiple clients can transmit and receive data simultaneously, which increases the number of clients that can be supported over the air and improves overall performance. Early testing is promising, but the performance gains will be better understood as Wi-Fi vendors complete field testing and gain more experience with live deployments. Naturally, clients will have to support these features, and it is likely future devices will, but it will take time before there is a substantial number of Wi-Fi 6-certified clients. For example, Apple's iPhone hasn't been Wi-Fi 6-certified by the Wi-Fi Alliance, which will impact iPhone and iPad users' ability to take advantage of Wi-Fi 6 early on.
The efficiency gains will be in large part due to the AP scheduler. Wi-Fi clients aren't involved in the airtime allocation other than responding to AP queries about status. The AP implements the scheduler because it knows the status of all the connected devices and their priorities, and is responsible for sorting out the competing client demands and allocating spectrum to each client. Scheduling algorithms are proprietary intellectual property that wireless networking vendors differentiate on, and there will be lots of differentiation as machine learning algorithms are developed to adapt to changing conditions in the airspace. Since the clients are not involved with scheduling, Wi-Fi vendors have a great deal of latitude in adding features and capabilities to the scheduler while still being standards-compliant and interoperating with any Wi-Fi client. Cellular Radio Access Network vendors will claim that scheduling is difficult, and that they have decades of experience building, testing and refining scheduling algorithms to eke out every bit of efficiency from the airwaves, which Wi-Fi vendors have yet to go through.
Vendors are already innovating with scheduling. Aruba's recent announcement on Air Slice is one such innovation, where the types of applications can be factored into the prioritization process and influence the resulting schedule. For example, rather than basing priorities on just the wireless client, Air Slice can identify applications in real time, such as VoIP, and schedule airtime to clients to maintain proper capacity, latency and jitter, resulting in better voice quality for VoIP clients. Expect Aruba to enhance this capability and for other vendors to follow suit with similar technical strategies.
Private 5G using the CBRS spectrum provides the much-needed spectrum management that Wi-Fi lacks. There are multiple FCC-authorized SAS operators from Amdocs, CommScope, Federated Wireless, Google and Sony. It will be incumbent on the SAS operator to allocate spectrum use and ensure that GAA tiers don't interfere with any Incumbent and PAL spectrum use. The GAA tier doesn't offer a guarantee from interference from other GAA users, and this creates an opportunity for SAS operators to differentiate on spectrum management, relying on analysis of the spectrum via machine learning and artificial intelligence techniques to respond in real time to changing radio conditions. The assumption is that – with indoor use, the low level of CBRS transmission power combined with attenuation through floors and walls, and the benefits of distance between radios – co-channel interference could be mitigated. These assumptions need to be validated in the field. Other vendors not yet approved by the FCC for SAS operations are researching spectrum management capabilities and hope to bring them before the commission in the next 12-18 months.
Sharing is not caring
Due to the radio spectrum management introduced in Wi-Fi 6, Wi-Fi vendors are claiming that they now have a deterministic radio system that is nearly competitive with cellular, but that isn't quite accurate. Telecom spectrum is licensed for exclusive use by operators, which don't have to worry about the problems of multiple users or organizations accessing a shared spectrum. Wi-Fi 6 adds useful features and capabilities to spectrum management, but unless the radio network is isolated from all other radio networks – like in the countryside – there will still be contention for the airwaves, which will unavoidably impact all of the Wi-Fi 6 networks. BSS coloring in Wi-Fi attempts to mitigate interference from other systems, but it has yet to be proven in the field. Even so, such mechanisms won't help in high-density environments like multi-tenant office buildings and urban and suburban areas.
Additionally, cellular protocols use multiple radio channels for transmitting and receiving, allowing for bi-directional communication and a management channel, and can support a high number of devices (numbering in the thousands per installation) in a much larger area than Wi-Fi. One of the outcomes is that as more clients are added to a tower, performance stays relatively flat for all devices as the number enters the thousands. Wi-Fi devices suffer performance degradation rather quickly as more devices are added due to contention of the airwaves.
The use of the recently opened 6Ghz space for Wi-Fi may help alleviate some of the airspace problems for Wi-Fi APs that support Wi-Fi 6E, at least in outdoor deployments, where the FCC has mandated the use of an Automated Frequency Controller to managed and allocate spectrum.
In addressing these challenges and obstacles, Wi-Fi vendors will need to:
Define the scenarios where Wi-Fi is a good fit and where another technology like 5G would be better.
Demonstrate that Wi-Fi can provide deterministic networking that satisfies enterprise requirements.
Provide training and tooling for IT to effectively manage Wi-Fi networks and monitor performance.
Create best practices to use Wi-Fi 6 in dense environments.
Private 5G vendors will need to:
Show how the capital and operational costs compare with Wi-Fi and clearly state the benefits.
Create products and services that are tailored for enterprise IT workflows and processes.
Identify and cultivate integrator and VAR partners that can sell, deploy and support private 5G.
Streamline the 5G device provisioning and management process.