The End of the Road for WiMAX!

I guess this industry needs its feeding frenzies in order to function but it seems as though we don’t learn from one to the next and make the same mistakes time and time again.

Sprint announced the end of life for its WiMAX (Clearwire’s) network on November 6, 2015. WiMAX burst onto the scene in June 2001 and was to change the wireless world forever. It attracted a lot of smart people and companies including Intel, which became the WiMAX ambassador around the world. Intel even purchased spectrum in one or more countries and promoted WiMAX as a royalty free wireless technology either at the high end of the 3G speed curve or the low end of the 4G (not yet invented) curve.

Typical of other overhyped technologies, even though many major companies and a number of start-ups entered the WiMAX feeding frenzy, WiMAX fizzled. This particular feeding frenzy was set off because the existing commercial marketplace for wireless technology was well established and royalties had to be paid to make use of the popular technologies. In the days of GSM development, the companies involved traded patent licenses and it worked out, but latecomers had to pay to play in both the GSM and CDMA technology spaces. WiMAX enabled newcomers and even some existing wireless providers (such as Sprint) to deploy a 3G+ or almost 4G technology for no royalty fees or costs of market entry.

I was told by a senior executive at Intel that Intel would never (his word not mine) license CDMA from Qualcomm because Intel was afraid it would always be a release behind. So Intel jumped at WiMAX, and once it was onboard many smaller companies and startups decided WiMAX might be their opportunity for fame and fortune, most notably fortune. It is not clear how much money Intel sunk into WiMAX, buying spectrum in Sweden in 2008, installing the system in Taiwan and making it operational, and spending large amounts of money to promote WiMAX and to ensure it would become successful. WiMAX’s run was over a decade but was marred with smaller company failures and WiMAX systems being planned and cancelled but there were also some wins for WiMAX—one big one in Russia, and then Clearwire in the United States.

I started writing about WiMAX in 2001 and followed its path for a number of years prior to LTE making its way onto the scene and basically trumping WiMAX in all regards. My first WiMAX articles were skeptical of both the technology and the players. The technology was a modified point-to-point microwave technology that was pressed into mobile broadband service and the players, including Intel and a host of others, all wanted to play because WiMAX was royalty free and, therefore, they would not have to pay fees to Qualcomm, Broadcom, Ericsson, Motorola, and other wireless players.

Because of its microwave roots, WiMAX was designed for Time Division Duplex (TDD) networks as opposed to the state of the art 1G, 2G, and 3G Frequency Division Duplex (FDD). This was a limiting factor in where within the spectrum it could be deployed, but companies such as Alverion, which was a leader in WiMAX equipment and systems for a short period of time, believed WiMAX could be modified for use in FDD spectrum designations. I wrote an article about WiMAX and TDD/FDD in April 2007 when it was still going strong and WiMAX folks were confident it could become an important 4G standard.

WiMAX obviously created a feeding frenzy among those who felt they had been frozen out of the commercial wireless market and WiMAX would provide new companies and new entrants with the opportunity to compete in a market already dominated by large, well-funded communications companies. As mentioned, Intel took up the WiMAX banner, touting its success as part of the new order of wireless. It invested heavily in WiMAX and convinced a number of people to create new businesses to serve the WiMAX market.

Clearwire was born prior to WiMAX in 1998 but in 2003 Craig McCaw purchased Clearwire Technologies, Inc., parent company Clearwire Holdings. McCaw brought in a lot of major shareholders including Sprint, Comcast, Time Warner, Bright House Networks, Google and, of course, Intel. After this new group came together the hype really started flowing. WiMAX was to be a real challenge to the wireless establishment, changing the face of the wireless world forever, and Clearwire had a lot of spectrum in the 2.5-GHz band available for TDD WiMAX. (Though from its inception WiMAX was meant to evolve to run on FDD systems as GSM and CDMA did, to my knowledge no one ever deployed anything other than a TDD system.[1])

In 2008, Clearwire made a big deal about launching WiMAX into commercial service in Baltimore. I was one of those invited for a tour of the system and a drive around the city in Clearwire’s test van. I was given the entire tour including being treated to streaming video (a movie) as we moved from cell site to cell site. The speed of the network was pretty good but we were the only user on it at that point. Clearwire faced a number of issues in Baltimore with obtaining enough backhaul and the network going up and down multiple times, but it was launched commercially and other cities were lit up afterward.

It is somewhat ironic that in August 2012, four years later, Sprint made a really big deal out of launching its LTE system in Baltimore; its press release even referred to the city as the birthplace of WiMAX.

As with all new wireless technologies, those promoting WiMAX were promising data speeds far higher than they could deliver. Early press releases told of 30 to 40 Mbps data speeds and the WiMAX Forum claimed that by 2011 it would provide speeds up to 1 Gbps (Gigabit per second). Speeds offered in the commercial systems were not even close to these promised speeds. The tests we ran in Baltimore and elsewhere showed per user data speeds in the 1-3 Mbps range.

In addition to Mobile WiMAX, the entire WiMAX family of products was supposed to provide point-to-point backhaul that could augment fiber distribution, providing a wireless alternative to DSL and cable “last mile” broadband access and Internet connectivity. There was also talk of using it for Smart Grids and metering for the electric utilities. Cable operators, at least those involved in Clearwire and later Sprint, were touting WiMAX use for their “triple-play” bundle of video, Internet, and voice calling.

Clearwire had a lot of spectrum—120 MHz in most areas. This spectrum was obtained from the partners that came together to form the wireless version of Clearwire, and today Sprint owns this spectrum and is replacing WiMAX with TDD LTE as fast as it can. With Sprint’s spectrum holdings in the PCS or 1900-MHz band, its Nextel spectrum in the 800-MHz band, and its 2.5-GHz spectrum, Sprint controls more wireless broadband spectrum than any other wireless network in the United States. However, AT&T and Verizon have spectrum in the 700-MHz and 800-MHz bands, which gives them a real advantage when it comes to fewer cell sites required to cover a given area. Excluding the need for capacity, for every 700-MHz site built by AT&T or Verizon, Sprint (Clearwire during its WiMAX days) had to build out five to seven sites to achieve the same coverage.

There were never many devices available for WiMAX because the volumes were low. A number of companies built USB modems and some plug-in devices for notebooks, but it was not until 2010 that the first phone, made by HTC, became available and it was not followed by any more phones. In fact, in 2012 Sprint stated it would no longer offer any WiMAX devices even though the network was still in operation. With Sprint finally announcing the end of life of WiMAX within its network, more than decade of the WiMAX attempt to challenge the established wireless technologies and networks has ended.

In reality, the end has been coming for a long time. In 2009, Alcatel-Lucent moved its WiMAX push to LTE and in 2010, Cisco exited the WiMAX business. As early as 2008, I wrote several articles, some appearing in national wireless publications and some in my own TELL IT LIKE IT IS blog, about Intel’s upcoming exit strategy for its WiMAX business. These articles were prompted by the then VP of marketing for Intel stating multiple times that Intel believed WiMAX and LTE should become one and the same. After I wrote the articles I was told, in no uncertain terms, that Intel would “never” (that dangerous word again) exit the WiMAX business and that it was dedicated to its success. They pointed to the fact that Intel had recently invested yet another $1 Billion into Clearwire.

Well, it was just about two years later in 2010 when Intel announced it was pulling its support for the WiMAX program in Taiwan and was dismantling its WiMAX program office. In 2012, Intel, licking its WiMAX wounds after finally pulling the life support lines to what was left of its WiMAX business, turned around and announced its support for TD-LTE or TDD LTE that will be used in China and by Sprint in its 2.5-GHz (ex-WiMAX) spectrum. So far, little has been heard about Intel’s success with TD-LTE although the technology is now deployed and working well in both China and the United States. Perhaps the main reason for this silence is that Qualcomm was quick to start marketing chipsets capable of both FDD LTE and TDD LTE and quickly captured a large portion of the TD-LTE market.

Lessons Learned?

Over the years there have been many “feeding frenzies” when it comes to wireless technology. WiMAX was perhaps the most well-known of them and was responsible for the start and failure of countless businesses that thought they would ride Intel and Clearwire’s coattails to the bank. Instead they are either gone or have shifted course to support LTE. There were other feeding frenzies early in the wireless data evolution including IBM’s invention of a technology it called Cellular Digital Packet Data or CDPD. Craig McCaw picked up and ran with CDPD, trying to convince the cellular world that CDPD was the “be all, end all” of wireless data and that the existing nationwide wireless data networks, run by RAM Mobile Data and ARDIS (Motorola after it was built for IBM) were too slow and did not provide the same coverage. Again, many small startups joined in and to its credit CDPD did have some success. Verizon did a good job with CDPD for Public Safety organizations but I am willing to bet that not a single wireless network ever made a profit from CDPD.

Next up was Bluetooth, now a common communications tool, designed and invented by Ericsson. Named for a famous Swedish warlord, the technology was basically given to Intel and it had the analyst community in to talk about how it would build Bluetooth into every Intel chip from then on and how it would change the world of communications. It is interesting that Intel never marketed or sold a single Bluetooth product and it took the Bluetooth SIG many years to mold it into what it was touted to be.

And how could we forget Municipal Wi-Fi—unlicensed spectrum, access points on light posts in a city, and mesh networks? Cities jumped at the chance to stick it to the network operators, providing their citizens with free or almost free Internet connectivity everywhere in a city (all of this after Metricom’s metro networks failed). Then reality set in—it does not work, it is very expensive, and there is no sound business model. Muni-Wi-Fi failed, was tried again, failed again, and still seems to be living on to fail again.

History is closing the book on WiMAX and LTE is the admitted world standard, but wait! There is another feeding frenzy on the horizon! 5G is coming! There are 5G conferences already, before there is even a single approved definition of what 5G is or will be. I guess this industry needs its feeding frenzies in order to function but it seems as though we don’t learn from one to the next and make the same mistakes time and time again.

This is what happens when you have a bunch of marketing types designing the next technology and defining for customers what it will be even before the engineers figure out what is possible when the laws of physics are applied.

Andrew M. Seybold