Wi-Fi Direct has just been announced by the Wi-Fi Alliance. It is a mechanism to add a soft access point in one device so that two different products can form a direct point to point secure wireless connection. Supposedly it will be easy to set up and relatively fast. The objective is to avoid having to configure a consumer device with all of the info required to use a standard WiFi access point and enable a somewhat simpler point to point file transfer.
The cartoon video on YouTube from the Wi-Fi alliance demonstrates a few of many potential use cases. Transfer of photos from one camera to a tablet or laptop, transferring photos from a camera to a printing kiosk, handheld to handheld gaming, and handheld to handheld biz card transfers. All of these applications are focused on file transfer oriented use cases (NOTE: they don’t involve streaming media, which is not one of Wi-Fi’s strengths).
But be careful . . . The applications cited, as well as other cable replacement applications, give Wi-Fi Direct the edge in essentially killing off any aspirations the Bluetooth SIG may have had in pursuing higher speed data transfers in the 2.4GHz spectrum and probably is the coup de grace for Bluetooth 3.0. I expect Wi-Fi Direct will find its way into some interesting convenience applications. However, it has some restrictions and future ramifications which are not obvious.
Wi-Fi Direct is likely to be limited to use in the already heavily congested 2.4GHz band. It seems that almost everything wireless in your house communicates on these same bands and there is only so much bandwidth to go around. In addition, these random data transfers will certainly mess up corporate Wi-Fi networks with interference. It should work fine in most suburban situations but most likely will have problems in apartment complexes and in business settings. One application it most certainly will tromp on is Intel’s WiDi wireless display which uses the same spectrum.
Interestingly, we are unlikely to see Wi-Fi Direct being used in the 5GHz; wind shear radars used at all of the world's major airports operate in this frequency band. In the US, the FCC has stringent rules about 5GHz since they don’t want to see a rental car bus full of “Wi-Fi Direct beaming business people” causing glitches on the radar screens at airports – so the FCC is being responsibly aggressive here. The esoterica of DFS (dynamic frequency selection) and avoiding radar interference is beyond the scope of this blog, but suffice it to say the only safe place for most manufacturers to intro Wi-Fi Direct will be in the 2.4GHz band that as we have noted is already saturated.
In the end what this all means is that for products that already have a Wi-Fi radio and some form of user interface you can expect Wi-Fi Direct data transfers to become enabled over the next few years. Smartphone connections to HDTV is one that comes to mind. While both Smartphones and cameras have integrated Wi-Fi technology today battery drain when using it has become of significant concern to users. OEMs will be very careful in the rollout because the connection in most cases will not be suitable for streaming – but rather bulk file transfers. Products where a user would expect video streaming may not be enabled since it won’t work well. Further, in many cases these new applications will not have a parallel application that currently requires a cable. As a result, there is no consumer transference from a current connectivity application to the more convenient wireless applications. Therefore, these new applications, though interesting, will take longer gain the attention of consumers and to penetrate into mainstream consumer market.
In the meantime, Wi-Fi Direct is a step closer to a future “life without wires”. However trying to use an already overcrowded 2.4GHz spectrum will undoubtedly cause a host of new problems as well as severely limiting the users bandwidth (and experience). Expect this new added convenience to likely cause performance issues if you own a Intel WiDi video system and to slow down your laptop internet access to your Wi-Fi access point. By the way, since I travel a lot, let’s hope the FCC keeps up its diligence to keep Wi-Fi Direct away from 5GHz TDWRs and from causing havoc with our airports wind shear radar. Naturally, as consumers try Wi-Fi Direct, many will like the convenience, but won’t like the low throughput, battery drain and unpredictable performance caused by 2.4GHz radio interference. To solve these problems manufacturers will turn to yet another radio – UWB. UWB is very small, low power, high throughput and was architected from its inception to do high speed point to point data transfers without suffering from Wi-Fi interference for both data files and streaming video.”
 Due to increasing interference being seen by radar operators at US airports, the FCC conducted testing and found a myriad of potential alarming interference issues. In a letter dated July 27th of 2010, the FCC commissioner issued a memorandum on the topic of 5GHz Wi-Fi interference with TDWRs (Terminal Doppler Weather Radar) asking operators to remove 5GHz equipment within 35m or line of site of airports. Eric Broockman is Founder and CEO of Alereon.
What WIFI-Direct looks a whole lot like is a solution desperately searching for a real problem. One big deal thing that has been completely ignored is security, which may not matter to gamers, but those downloading photos or other files may not wish them to be available to the general public. Encryption is a burden that uses battery power and often slows the transmission process, so it is not likely to be included. Besides that, it will be quite funny when the entire 2.4 GHz spectrum becomes filled and only the strongest signal gets through. That will indeed be an interesting scene.
So how will Wifi-direct devices pair?
on Bluetooth most of times is a simple PIN number exchange, but when I do a Wi-Fi connection to a hot spot, many times the WEP key isn't that easy isn't it... so... isn't the code going to be a little more difficult than current bluetooth?
The points you are bringing are very valid. But I am not sure handset manufacturers are ready to integrate another complex radio such as UWB.
I think Bluetooth 3.0+HS is a better candidate than Wi-Fi Direct in a sense that it is benefiting from all the classic Bluetooth infrastructure that people are now familiar with, like pairing mechanism and Bluetooth profiles that ensure interroperability at application level. But I agree this will not solve the issues you are highlighting. Some improvement of the Bluetooth 3.0+HS could make it, with for instance a clever Wi-Fi channel selection that minimizes the interferences with surrounding AP.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.