I recently got a new Wireless-N router (based on the Draft 2.0 of the 802.11n spec.) to replace my old, but trusty WRT-54GL, an 802.11g router. I bought this router just over a year ago to replace another 802.11g router that was acting up. I specifically bought the WRT-54GL (as opposed to the regular WRT-54G) because I knew that I’d eventually want to flash it with third-party firmware. (The WRT-54GL runs on Linux and has better hardware than later versions of the WRT-54G)

I recently got my Xbox (with XBMC) setup in my living room. This placed it far away from the wireless router in my apartment and not wanting to run an unsightly CAT-5 cable around my place, I decided that I’d need a wireless bridge since the original Xbox does not have a built-in wireless adapter. (Nor does the 360, I believe). A quick look around the web showed that most dedicated wireless bridges were quite expensive and of dubious quality. Since I already had the spare WRT-54GL, I decided to flash it with DD-WRT, one of the most popular third-party firmware replacements, and see how it would function as a wireless bridge for my XBMC setup.

Bridging the gap

A wireless bridge (depending on your definition) is a device that allows regular wired clients access to the network over a wireless connection. This was exactly what I needed for my XBMC-equipped Xbox. This would give XBMC access to the network and allow me to play content/videos stored on my PC using XBMC, effectively turning it into a streaming box. Doing a little research, I found out that wireless bridge capabilities were one of the many features that the DD-WRT firmware added to the WRT-54GL.

Mind you, DD-WRT doesn’t only work with the Linksys WRT-54G/GL. A full list of supported devices is available on their site, and the list is quite long. However, I believe the WRT-54G was the first, after people discovered that initial versions of the WRT-54G were running Linux, thus requiring Linksys to make the source code available and spawning a bevy of third-party development for the device. (Some later versions of the WRT-54G unfortunately do not support the full version of DD-WRT since the newer WRT-54Gs are now being shipped with VxWorks, another embedded OS, instead of Linux and thus have different hardware configurations)

Getting it to run

DD-WRT has a feature list so long that covering it is outside the scope of this article. Instead, I’ll focus on how I set it up as a wireless bridge to provide wireless access to the network for wired clients. The DD-WRT wiki has an installation guide, and while it gets the job done, it’s unfortunately very convoluted, confusing and contradictory, on account of it being written wiki-style by many authors with evidently no editing whatsoever.

Basically, here’s what I did to get my WRT-54GL flashed to DD-WRT from the regular Linksys firmware: (As adapted from the installation guide, which you may want to follow as well) If you do not have another router to use while you’re flashing, you’ll definitely want to read through the instructions thoroughly before doing anything to make sure you have all the required materials, instructions and so forth, since if anything goes wrong, you won’t have an Internet connection!

1. Make sure you are on a wired connection to the router.
2. Reset it to factory defaults.
3. Download the latest stable MINI version of DD-WRT. Flash using the “generic” bin over the web console. The guide recommends using Internet Explorer instead of Firefox and while those instructions may be outdated, I followed them anyways.
4. If, after flashing, you cannot login with the default root/admin (which happened with me), do a hard reset of the router. This is accomplished by holding down reset and plugging in the router while continuing to hold reset for 30 seconds. (You may just have to hold reset for 30 seconds without the plug-in procedure) This should reset the login/password back to the normal defaults
5. Now, get the latest stable STANDARD version of DD-WRT. Flash again using this version’s “generic” bin over the web console. This adds more features over the mini version, which apparently you must first flash with when changing from the default Linksys firmware.

This procedure worked for me, but I can’t be sure it’ll work the same for every router out there.

Getting the wireless bridge connected

The documentation for this over at the DD-WRT site, while long, is less confusing than the installation instructions. I basically followed them to the letter, and within a few minutes, the WRT-54GL was online and connected over a WPA-PSK secured wireless connection. I definitely recommend some form of WPA over WEP, as WEP is broken and cannot be considered secure for anything.

The guide indicates that WPA2 does not work correctly for the current version of DD-WRT (2.3 SP2 as of this writing), though it is an option in the settings for a wireless bridge. I have not tried to get WPA2 to work myself, but reports from others seem to indicate that it does not work. If and when I do try, I’ll update this article with my results.

Well, that’s about it. I’m now happily streaming video to my XBMC setup, and so far, the wireless connection has been rock solid, without a single drop-out, freeze-up or slowdown. In fact, it feels as zippy as a wired connection! This is what wireless networking should be like and I’m glad it’s finally becoming a reality.

Despite all the coverage that Apple’s new iPhone is getting, there is still innovation going on at other cellular providers. Last week, T-Mobile announced their new “HotSpot @Home” service. For an additional $10 a month, you can use your WiFi-equipped cellphone as a VoIP device. When you’re in range of a WiFi access point, the cellphone will use the Internet to make VoIP calls (instead of the cellular system) – and the minutes used here won’t count towards your monthly plan, giving you essentially unlimited usage time when you make calls this way. Your phone still retains all its original functionality, such as call hold and caller ID, so your phone isn’t crippled.

The best part, however, is the seamless integration. As you move out of range of a hotspot, the phone will automatically hand off to T-Mobile’s cellular network, in much the same way as hand off occurs between adjacent cellphone towers when you’re moving. This is not only beneficial, but a prime example of the convergence of wireless networks that many people believe is going to play a bigger and bigger role in the future. Kudos to T-Mobile for offering innovation in this area, instead of just a new phone with a lot of fancy features.

iPhone blowback

When the iPhone was announced, the hype was almost immediate. For the most part, it’s lived up to its expectations, but two issues identified before launch seem to have no resolution. Firstly, it can only be used with AT&T. Secondly, and perhaps more importantly, was the lack of an SDK for the iPhone. Many people lamented this decision, and wondered if Safari and web-apps would be able to replace the flexibility offered by a true SDK. Despite worries of data usage costs (for the eventual non-US users), there have already been a plethora of iPhone apps released.

The real reason, some say, for the lack of an SDK is perhaps Apple’s (and AT&T/Cingular’s) worry about a VoIP application being developed for the iPhone that has the potential to severely cut into their income. After all, phone calls over the Internet wouldn’t count as regular minutes. Additionally, AT&T is known to harbour fears of companies that “use their” pipes, especially when they’re offering services that could cut into their bottom line. The result of this is a stifling of innovation.

Convergence of networks

As I mentioned before, wireless convergence is a key component of next-generation networks. The idea of separate devices operating on separate wireless networks is really an archaic one, held in place partly by companies unwilling to compete in a world of open, convergence networks. Convergence is not only good for the end-user/customer, but also good for the companies that provide the service. Customers get the best of both worlds: Cheap calls while you’re localized, and you still have wide coverage when you’re not.

There is also a clear need for a service like this. Many people already use Skype to make cheap or free VoIP calls. There are even WiFi Skype Handsets available to use with the service. T-Mobile’s new service basically integrates that functionality with a cellphone, and thus has the potential to take over that market with a more useful device.

The benefits for companies are perhaps even greater. Take T-Mobile’s HotSpot @Home service as an example. Because the phone will use a WiFi network to make VoIP calls when one is available, this not only provides better coverage but also decreases the strain on T-Mobile’s own infrastructure. Since many people have WiFi at home and at work, there will be less users on T-Mobile’s cellular system, decreasing its load and increasing its ability to support more customers. True, T-Mobile still has to have infrastructure on their own end for translating VoIP calls to landlines and vice-versa, but the benefits of less cellular load clearly outweigh this.

Which leads me to question why they’re even charging an additional $10 a month for this service. Since it’s beneficial to them, shouldn’t they be offering this for free? Indeed, if more cellular providers started offering these kinds services, they would be a true force against Skype. They wouldn’t need to fear it, or try to inhibit other VoIP providers, if they themselves could come up with an improved solution like this. They need not charge more, since the increased usefulness will attract more customers, and thus increase their profits naturally. But, perhaps I’m just nitpicking here, since no other company offers a service like T-Mobile’s, and is a true example of using innovation to compete, instead of stifling it.

Let’s hope this sort of service comes to Canada. If it does, I might truly have no need for a traditional land line.