Seven common misconceptions about the 802.11ax wireless LAN standard

For the 802.11ax wireless LAN standard, which is very likely to be standardized in 2019, now may be the beginning of market warming up. Many manufacturers have published their own product technology white papers and even product roadmaps, and some manufacturers can't wait to "launch" related products.

However, as wireless network engineers, we should be aware that technology is for users' business services. The intention of launching each new generation of technology, standards and related products is good, trying to solve a series of problems that users are facing now. Unfortunately, the ideal is full, but the reality is very thin. The new era faces new problems. New technologies and related products may not completely solve the pain points faced by users, or solving related problems requires involving a complex industry ecosystem, which is a mountain that we need to overcome in today's era.

[[232728]]

This article will not go into the technical details of 802.11ax, as I wrote a special article about it in July last year. What I will talk about next is what 802.11ax can do under what conditions, and its limitations (what it cannot do).

Since the 802.11n era, the spectrum that Wi-Fi relies on has been basically developed and exhausted. Unless more spectrum resources are opened up, it is difficult for the 802.11 protocol family to make much progress in the way of making the road wider and wider. Therefore, 802.11ac began to work on how to increase system performance and capacity, such as multi-user multiple input multiple output (MU-MIMO) and denser modulation methods, but overall, 802.11ac's efforts in this regard are not satisfactory. The industry has also recognized this. 802.11ax's series of enhancements to the 802.11 physical layer (PHY) will help us use limited channel space more effectively to achieve higher efficiency.

However, these innovations require certain conditions to be realized, and are by no means the ultimate solution to all the problems we face. The following are some of the misunderstandings about 802.11ax reflected in some of the questions that are frequently asked recently.

Myth 1: 802.11ax is a cure-all!

The truth: 802.11ax cannot solve the problems caused by incorrect planning and design

We have spent a lot of time explaining that planning and design are the top priorities for wireless networks, and adhering to correct planning and design is the first step towards successful deployment. Only with correct design and planning can good products and system features be brought to their full potential! No matter how good the products and technologies are, they cannot make up for design defects. Compromises in planning, design and product selection will open the door to failure.

In actual design and deployment, simply increasing the speed is not a panacea. Upgrading the 802.11 technical standard itself cannot solve the various mistakes we made during the planning, design and deployment stages.

This is because in today's actual deployment environment of wireless LAN, the goal is high-capacity access, and the distance between wireless access points is much closer than the previous coverage-oriented method. The limited spectrum resources will result in low channel reuse efficiency. The first task of reasonable planning and design is to reduce medium competition and reduce radio frequency interference, that is, efficient channel reuse to minimize air interface sharing between wireless access points. This is the premise, and then it is to increase the efficiency of air interface utilization within the coverage cell of a single channel, that is, to establish a higher data connection rate between the client and the wireless access point, which is also the continuous improvement of the 802.11 standard. If the premise cannot be achieved, simply improving the latter is meaningless!

If you can upgrade your clients to support 802.11ax as soon as possible, then the wireless network infrastructure based on 802.11.ax may, perhaps, increase the capacity of your network. The reason why I say "may" and "perhaps" is that as an engineer with 12 years of wireless network experience, I have seen that more than 80% of Wi-Fi network designs and deployments have more or less design planning and deployment problems, so that there is no "magic" technology or product that can help them correct these problems. This is not an exaggeration, many of them are lessons learned and painful experiences of users.

Myth 2: 802.11ax can solve the problems facing the 2.4GHz spectrum

The truth: 802.11ax is not a cure for the problems facing the 2.4GHz spectrum

The 2.4GHz spectrum is dead! Although 802.11ac and 802.11ax wireless access points still support the 2.4GHz spectrum, this does not mean that these protocols can do more in the 2.4GHz spectrum. Apart from the numerous Wi-Fi and non-Wi-Fi devices and interference sources, the core problem of the 2.4GHz spectrum is that there are only 3 non-overlapping channels available. Due to the large number of legacy clients, 802.11ax will not help you at all in 2.4GHz.

Of course, from another perspective, if a large number of enterprise clients use 5GHz spectrum to connect, the dilemma of 2.4GHz spectrum will inevitably be alleviated. However, this has nothing to do with the 802.11ax protocol itself, and the same effect can be achieved using 802.11ac.

Myth 3: 802.11ax can easily help you improve the performance or coverage of traditional (11a/b/g/n/ac) clients

The truth: 802.11ax is unlikely to effectively improve the performance of traditional clients

While you don’t have to worry about interoperability between 802.11ax APs and legacy protocol clients, it is difficult for 802.11ax APs to cost-effectively improve performance or coverage for legacy (11a/b/g/n/ac) clients.

Many people would say that since 802.11ax wireless access points have more antennas and MIMO processing capabilities (8×8:8), does that mean that the wireless access points will be more sensitive to wireless client signals? Will the ability to recover client signals be better? This is indeed the case, but chip/network manufacturers will not implement it this way. Why? Because too much is as bad as too little. Any commercial wireless access point that exceeds 4×4:4 is a waste, difficult to implement and costly, and the efforts to achieve additional uplink gain by increasing sensitivity ultimately have little improvement in sensitivity and reliability. This is also the reason why the second-generation 802.11ac did not commercialize 8×8:8 (also specified in the 802.11ac standard) wireless access point products.

Of course, it is not ruled out that chip/network manufacturers may launch related products. In this case, I believe the cost must be high, and this part of the cost will inevitably be shared by users.

Misconception 4: Only talking about the effect, not the prerequisites for realization

Fact: 802.11ax requires wireless network infrastructure and wireless clients to work together to achieve its goals

In simple terms, 802.11ax wireless clients are to 802.11ax wireless access points what 802.11ac wireless clients are to 802.11ac wireless access points. Without the support of established clients, wireless network infrastructure is unlikely to be of much use.

For example, if you use an 802.11n client to connect to an 802.11ac access point, or a first-generation 802.11ac client to connect to a second-generation 802.11ac access point, the access point must be downgraded to be compatible with these clients, and the performance improvement you get is basically negligible.

In real-world deployments, it is difficult to ensure that wireless access points work in an ideal and pure mode due to the life cycle of wireless clients. Even with the explosive growth of wireless network terminals today, and despite the continuous emergence of second-generation 802.11ac clients, we have only achieved the mass popularization of first-generation 802.11ac wireless clients.

Below are two categories of technical highlights of 802.11ax, which are also the key to improving the performance of wireless networks. Red fonts indicate that client cooperation is required to achieve this.

1. Physical layer enhancement and efficiency, mainly including:

• Orthogonal Frequency Division Multiple Access (OFDMA) in both uplink and downlink directions - requires client cooperation
• Multi-User Multiple Input Multiple Output (MU-MIMO) in both uplink and downlink directions - Downlink MU-MIMO requires 802.11ac Wave 2 clients, and uplink MU-MIMO requires 802.11ax wireless clients
• Uplink resource scheduling - requires client cooperation
• Up to 8 transmit antennas and 8 receive antennas, 8 spatial streams
• Higher modulation, 1024QAM - requires client cooperation

2. MAC layer enhancement and efficiency, mainly including:

• BSS Coloring - requires client cooperation
• Dual NAV mechanism - requires client cooperation
• Target Wakeup Time (TWT) - requires client cooperation

It can be seen that without the cooperation of the client, the effect achieved by the 802.11ac wireless client connecting to the 802.11ax wireless access point is roughly the same as that achieved by connecting to the 802.11ac wireless access point. Even in a mixed environment of 802.11ax and 802.11ac wireless clients, the final system energy efficiency is determined by the number of 802.11ax wireless clients.

Myth 5: 802.11ax is a switched wireless network

The truth: 802.11ax does not implement Wi-Fi switching capabilities

Wi-Fi (802.11) still relies on a shared medium - or half-duplex air medium - to transmit electromagnetic waves. The medium is still occupied by carrier sense multiple access/collision avoidance (CSMA/CA). We often use Ethernet hubs as an analogy.

The MU-MIMO mechanism introduced by 802.11ac (802.11ac is downlink [DL], 802.11ax also introduces uplink [UL]) does not realize the "exchange" capability based on the shared air medium. It is just that when the wireless endpoint wins the opportunity to occupy the medium, it can send to multiple clients on the downlink at the same time, or enable multiple clients to send using the uplink at the same time. This access mechanism is more efficient than sending one by one. However, the MU-MIMO of the 802.11ax uplink requires that the client must be an 802.11ax client, and traditional clients cannot participate.

Myth 6: 802.11ax will deliver all innovations at once

Fact: 802.11ax won’t deliver all innovations at once

Just as the 802.11ac protocol was introduced to the market in two waves, 802.11ax will also enter the market in the same form. The first generation of 802.11ax chipsets will not have the following features:

• MCS 10 and 11 (1024QAM) modulation
• 8 spatial streams
• BSS coloring mechanism
• Uplink Multi-User-MIMO

Myth 7: 802.11ax is here, buy it now

The truth: Enterprise-grade 802.11ax products are not yet mature

Both 802.11ax draft 1.0 and 2.0 versions have not been voted through. Currently, draft 3.0 is under discussion, and formal standardization is expected to be completed in the second half of 2019. Of course, the expected commercial products will be based on draft 2.0 or 3.0, and are expected to support the final standard through software upgrades.

Many advanced features of 802.11ax wireless networks require the participation of 802.11ax clients to be realized. In the short term of 2-3 years, there is no sign of the popularization of 802.11ax clients.

The so-called 802.11ax wireless access point products launched by some manufacturers on the market are actually based on early draft version chips, and cannot be upgraded to even the draft version through software.

To sum up, looking at an innovation requires a comprehensive perspective, and any blind following is not advisable.

It is rational to choose the final technology based on your actual application scenario.