The industry is well into the 802.11ac Wave 1 cycle now, with almost every vendor having product available. Now is the time many of us industry watchers start looking ahead at what’s next and for Wi-Fi that’s 802.11ac Wave 2. Wave 2 brings more bandwidth, in theory a little over 3.5 Gig; a remarkable speed for Wi-Fi.
The other innovation that Wave 2 brings is something called “beam forming.” Without going into a lot of technical details, beam forming concentrates the Wi-Fi signal and aims it directly at the target, as opposed to traditional wireless broadcasts, which distribute a signal to a wide area and with the hope that it’ll reach the device. So traditional Wi-Fi coverage looks like a circle, whereas beam forming looks like a star.
For those familiar with the Wi-Fi vendor Ruckus, the company has a proprietary version of beam forming called BeamFlex that has been a significant point of differentiation for Ruckus since the company launched.
Ruckus implemented BeamFlex and it worked well as a competitive differentiator, as there wasn’t a defined standard for beam forming available prior to 802.11ac. However, in 802.11 Wave2, beam forming will be a standard feature on the chips and all vendors will offer it as something that’s table stakes. In reality, Broadcom has already implemented single user beam forming into its silicon, giving all the Broadcom-based Wi-Fi vendors that capability today. Wave 2 is when multi-user beam forming becomes a standard that will close the gap even more.
The standardization of beam forming is a huge threat to Ruckus, as BeamFlex no longer differentiates the company from anyone else. Once school of thought I have heard is that because BeamFlex is proprietary, Ruckus may be able to stay ahead of the curve and make BeamFlex better than standards-based beam forming. However, I think the opposite is true. Since beam forming is a standard, it can be implemented in both the AP and the client, removing any advantage of BeamFlex possibly causing BeamFlex to underperform beam forming.
This scenario has actually happened to Meru Networks in the past. In the early days of Meru, mainstream Wi-Fi was 802.11 b/g running at 2.4 Ghz. Meru had some cool technology that it promoted as its “single channel architecture” that solved a couple of big problems. The first was that roaming from AP to AP was a big problem. Also, the availability of only three spectrally independent channels made pervasive Wi-Fi very difficult to plan without co-channel interference. This made applications like voice-over-Wi-Fi almost impossible to implement for everyone but Meru.
Meru solved this problem by coordinating access points and implementing proprietary signaling that didn’t violate the Wi-Fi certification but coordinated traffic on a packet-by-packet basis while avoiding collisions and minimizing co-channel interference from neighboring clients and APs. The signaling was interpreted by the client correctly, but it wasn’t something that the client actively participated in. In many ways, single channel was Meru’s “BeamFlex.”
Over time, though, the standards evolved to deliver pervasive Wi-Fi coverage and mobility, so handoffs were no longer a problem. The 802.11n chips were released and the standardized spectral re-use dramatically took away the throughput advantages that Meru had enjoyed. The improvements also operated in the 5 Ghz band, allowing for more mainstream use in clients, which further improved co-channel interference issues.
Another point that applied to Meru, and which is now true for Ruckus, is that they fell behind technologically. At Meru, with a/b/g products, the company had to design its own FPGAs because it needed specific MAC functionality. This inevitably led to Meru being late with product, as the chip manufactures made the standards-based chips far faster than a small company like Meru could. This is why the g product was released later than most other vendor products. I remember a conversation with a person I’ll leave nameless years ago that called building your own chip “fools gold,” as they may give you a slight advantage at first, but unless you’re the size of Cisco, it’s impossible to stay ahead of companies like Broadcom and TI.
The Ruckus proprietary antenna technology seems to have suffered the same fate. The Broadcom-based 802.11ac products from vendors such as Aruba and Aerohive have been available now for about a year, but Ruckus just started shipping its AC products last quarter, causing it to miss many of the early adopters. I do believe that the Ruckus antenna array is among the best in the industry, but it’s not so much better that customers should wait for other advanced features.
This actually brings up a larger issue that buyers should be aware of - vendors that differentiate themselves on RF will not have a product that can remain differentiated for very long. Aruba has managed to make security a differentiator for years, and Aerohive has focused on simplifying Wi-Fi deployments through the cloud and controller-less solutions. These are differentiators that can remain advantages as Wi-Fi evolves, whereas RF differentiation is likely to be short-lived.