Qualcomm sees 200m WiFi AP revolution in US alone

We suspect there is going to be both a patents battle in WiFi mesh software and that the reason this is important is that operators around the world will move lock, stock and barrel to multiple WiFi Access Points in each home, potentially doubling or even trebling the number of devices in next generation WiFi.

A side effect of this is likely to be more robust in-home WiFi, a dramatic reduction in retail WiFi devices and a reducing need for wired backhaul in most homes. And it all comes down to technical improvements we have been writing about here in Faultline Online reporter for the past 6 or 7 years.

We managed to conclude all of this from one telephone conversation this week with Irvind Ghai, the Qualcomm Atheros, VP Product Management, in charge of WiFi mesh.

First let’s take us back about 8 months, when Comcast came out with a surprise announcement at CES and said that by the end of 2017 it would offer software which would allow both the consumer and a remote help desk to control how WiFi is operating and it will roll this out to all of its 15 million Xfinity customers. Back in May it followed this up by saying it had signed a deal with US firm Plume (née Wildfire.Exchange) and Comcast took a shareholding in the company.

Ghai revealed that all those Plume devices can have Atheros chips inside them, not too surprising since it was started by ex-Atheros people, but essentially Plume is supposed to be hardware agnostic. But more to the point, they will work with a Comcast X3 router which also has an Atheros chip inside it, so as Plume Adapts its technology all those chips at Comcast Xfinity could come out of Atheros.

At 15 million homes, and 3 Access Point (APs) per home, that should be some 45 million chips – in a year. In a country where there are only 118 million homes, that’s a lot of chips.

Ghai also said that he has two more tier 1 accounts which he cannot reveal as yet, which will show their hands this year, and we have already seen Broadcom and Quantenna licensing software very akin to Plume mesh software from Turkish firm AirTies, and have suggested that this is to retain supply deals with someone big, possibly AT&T. AirTies has already licensed Sky in the UK, Germany and Italy with its WiFi Mesh technology which even allows for powerline as an alternative backhaul, where 5 GHz WiFi signals are not clear enough.

So that’s around 4 of the US majors about to go to market with very aggressive and very similar services covering as many as 74 million homes over the next few years and at 3 APs per home using up 222 million APs. Potentially this will also improve performance for any Homespot based WiFi network, like the one that Comcast will use to support its new Xfinity mobile service.

Fundamentally this next generation all relies on mesh, and the key ingredient of any mesh system is that each AP knows what is going on with the other APs in the mesh. They know if they change channel and have access to channel state information stored for neighboring channels, in case they have to jump to a spare channel, which is less congested. For instance they know if they have a neighboring AP from next door interfering and reducing their performance level even if they cannot witness that directly, but get it from another mesh node. Until now most retail products which have targeted improving your home WiFi in the US, have focused on selling extender products which work in exactly the same way, except that they do not share information, they merely pass on the signal.

This makes the WiFi vulnerable to local interference and unable to offer client steering and band steering features. The emergence of Plume as a retail product is meant to do away with this. It is a genuine mesh product which has controlling software in the cloud, which takes all of this information that the mesh has, and takes clever policy decisions on how your WiFi should work. It uses client steering, which is where the network decides which AP is best for a particular client, and band steering, where it decides whether 2.4 GH or 5 MHz is best for the client. It is tough to work out how much of this is native on the Qualcomm chips and how much is managed by the Plume software.

But Plume has the job of integrating its tiny devices which are little more than electrical plugs with radios and antennas attached, which take an existing WiFi signal and spread it around the home in an intelligent mesh. With the Comcast service they have to get the existing router to become part of the mesh, and this can then report back to a cloud server with all the home network’s data. It’s really a hybrid of what Plume was originally trying to do.

This is a fantastic improvement for an operator’s help desk, which can avoid a huge percentage of truck rolls, because it knows by using that cloud data, what is going on with the home WiFi network. It can then use client steering and other facilities like cutting off a troublesome device or putting one device WiFi to sleep, to stop whatever is wrong with the network. The next step is to anticipate these changes and do them automatically and also to put them in the hands of the customer, who can control things through an app, such as putting their children’s devices’ WiFi connections to sleep, when they the child is supposed to sleep. Add stuff like a cloud based parental control and you have a tons of reasons to take your WiFi from your operator, not from retail.

Ghai didn’t like our conclusion that this would lead to more operator controlled WiFi, but conceded that many of the reasons people first went to a retail store and swapped out their home router, is because of problems with WiFi penetration around the home, and out of a desire to have some control over it, when it went wrong.

AirTies has already landed 3 tier 2 deals in the US and Israeli rival Celeno has also claimed that it has a tier 1 account coming in the US.

It was then that Ghai referenced one of our articles citing both Broadcom and Quantenna licensing the software from AirTies, and said that Plume could do similar licenses.

We looked at the patents which are now attributed to Plume, having begun life as Wildfire.Exchange and saw that it has patents dating back to 2012 relating to things such as WiFi authentication from a cloud AAA server; Dynamic channel selection algorithms; hidden nodes detection; optimization of Wi-Fi access points – all things which are precisely what we have seen coming out of AirTies in the 2011 time frame. If they ever go to court on who invented what first, it looks clear that many of the Plume patents would not survive and since they are both suddenly falling back into a licensing model with the “big” players in the US market – people who use equipment suppliers who also dominate outside the US market – then with the world’s WiFi ready to enter a new generation, there will be everything to play for and that’s usually when companies do one of two things – they either license everything from everyone, or they go to court in patent claims. It will be interesting to see which way the world goes in this instance.

Plume originally came to market with what it called a “router-killer” which intelligently balanced home bandwidth between devices and offered plug-in pods for $49 each, which extend Wi-Fi throughout the home. But someone needs to operate the DOCSIS or ADSL modem to bring the broadband line into the home, and then that signal needs to be delivered into the Plume devices, usually by plugging the two together over an ethernet cable.

Comcast is integrating its xFi service into a unified experience and will this year release its own xFi pods (from Plume) which can pair with Comcast’s gateways to mesh WiFi throughout the home.

In the meantime we’d like to know whatever happened to Qualcomm’s own Streamboost (a cloud based QoS system) and its WiSON (WiFi Self Organizing Network) which tapped similar capabilities to those it is now boasting major US operators are about to bring out using Plume.

WiSON was supposed to deliver corner-to-corner WiFi around the home, as well as automated management and Netgear, Linksys D-Link, TP-Link ASUS and HiveSpot all said last year they are using it. It seems that the deals it does with operators put Qualcomm in competition with these suppliers. Ghai commented, “The retail supply chain knows it has to innovate and get to 802.11AX early, and bring in things like voice integration to stay ahead of operators.” Given that Comcast has 13 million voice remotes already, we cannot see a Comcast home being tempted by voice integration.