Sprint is to close down another network – having finally killed off iDEN last year, it will switch off Clearwire’s WiMAX systems on November 6 2015.
This marks the end of what now seems a small era, just a blip in the progress of the 3GPP juggernaut, and one that lasted less time than iDEN and had far less impact than CDMA. Back in 2008, that outcome wasn’t obvious though. It might have been clear that Intel would not fulfil its wildest dreams and see the WiMAX technology it sponsored taking precedence over LTE in the 4G era.
The arguments were logical – WiMAX was architecturally similar to LTE, built around OFDM and IP, but several years ahead; and it would support a more open ecosystem than LTE, new competition in the closed-shop world of mobile networks, and some of the more attractive IPR practices of the web/IT world.
And then WiMAX even succeeded in signing up a tier one mobile operator to its cause. Sprint was by far the most enthusiastic carrier about the technology, partly because it had a tradition of experimenting with left-field platforms for performance edge, partly because it needed to do something dramatic to narrow the gap with AT&T and Verizon. It had made one smart move, opening up to the MVNO model at an early stage, and one disastrous one, buying Nextel just as that firm’s once-loved iDEN services were on the wane.
Sprint’s next big idea was to harness the world’s first 4G technology and so gain a big headstart over its rivals as well as gaining global influence over the ecosystem as Verizon once had by backing the CDMA ‘pretender’. It launched its own WiMAX operation, Xohm, which it quickly merged into a joint venture with the US’s other big owner of 2.5GHz spectrum, Clearwire.
But perhaps inevitably, old habits and power bases reasserted themselves. The 3GPP community swung into action and accelerated LTE development so as to narrow the gap with WiMAX considerably – and importantly, the initial LTE work was focused on FDD spectrum, which most major cellcos used for 2G/3G, rather than WiMAX’s TDD. That helped get the operators and device makers comfortable that LTE was the safer bet and Sprint, with a few others such as Japan’s UQ, found itself somewhat isolated, with WiMAX starting to retreat to its roots in fixed and metro wireless.
Sprint’s old habits didn’t help either. The Clearwire JV started with grand visions of disrupting the US telecoms scene by following a new model, one focused heavily on wholesale – supporting a host of new service providers in an extension of Sprint’s MVNO deals. Google and Intel were sufficiently convinced of this brave new world that they invested in Clearwire, as did three cablecos. The prospect was of Clearwire supporting an alliance of the US’s third cellco plus three MSOs – a partnership with the scale, networks and content to compete with the quad plays emerging from the big two.
But it all fell apart. WiMAX proved a technology dead end; Google deserted Clearwire, as did the cablecos in time, for a new strategic alliance with Intel. Sprint did not have the courage of its convictions and increasingly treated Clearwire purely as a resource for its own retail services, and it had an increasingly acrimonious relationship with its JV until it finally acquired full control of it, from under the claws of Dish, last year.
Clearwire had already put a TD-LTE migration plan in place and Sprint accelerated that and made the 2.5GHz spectrum and networks a fully integrated part of its latest grand plans – the Spark LTE service, which will run across its three bands to deliver superfast 4G harnessing the plentiful 2.5GHz frequencies; and Network Vision, a multi-technology, multiband infrastructure which could host multiple operators’ networks and so resurrect the wholesale-focused idea once again.
Sprint said enterprise WiMAX customers would be given notice a year before the network is closed, and consumers will be notified six months in advance. It promises comparable devices at low or no cost to replace WiMAX equipment. The WiMAX network covers about one-third of the US population and has been losing customers, especially since Sprint itself stopped selling WiMAX handsets under its own brand. The service is still supporting some customers of its inhouse MVNOs, Virgin Mobile and Boost Mobile, as well as external MVNOs like FreedomPop, Ting and Karma, but all these providers are in the process of converting customers to LTE.
Sprint also said it had kicked off a “workforce reduction plan” to cut costs and become “more competitive in the marketplace”. It expects to record a charge of $160m for severance and related costs but did not specify how many jobs would be lost. They will come on top of 1,400 reductions made earlier this year.
Carrier aggregation advances:
Carrier aggregation is the most rapidly adopted element of LTE-Advanced and it is increasingly being used to augment FDD roll-outs with TDD spectrum, as Sprint will do.
Australia’s Optus last month launched commercial services boosted by carrier aggregation, in FDD/TDD spectrum. It has upgraded its networks in Sydney, Melbourne, Brisbane and Adelaide to support theoretical peak speeds of 220Mbps, by aggregating its existing 4G networks with its 2.3GHz spectrum.
That high frequency TDD spectrum was acquired with Unwired Australia in 2012, and like Sprint, Optus has been shutting down Unwired’s WiMAX systems, and converting to TD-LTE.
T-Mobile USA is deploying aggregation across its three FDD bands to boost its LTE speeds, working with its suppliers Nokia Networks and Ericsson. Their equipment can already support CA across the carrier’s 700MHz A Block, 1.7/2.1GHz AWS and 1.9GHz PCS spectrum. TMo uses the Nokia Flexi Multiradio 10 base stations for its 700MHz and AWS LTE sites and its LTE/GSM RF sharing platform for 1.9GHz, which is mainly being targeted for rural coverage. It also uses Ericsson RBS6000 base stations across its bands. The bulk of the LTE build-out so far has been in AWS, where TMo has historically run 3G.
Of course, software upgrades and supporting devices will be needed to commercialize the aggregation capability and TMo has not given a date for that, only saying it is “actively testing” triband CA. It is also adopting other LTE-Advanced features to bolster its performance, including 4×2 MIMO, which is being harnessed to improve cell edge performance and will go nationwide in time. Nokia has also deployed 4×4 MIMO in its LTE base stations.
AT&T Mobility has deployed carrier aggregation over its 700MHz and AWS frequencies while Sprint will start harnessing CA to add the capacity of its 2.5GHz TDD band to its 800MHz/1.9GHz FDD network by year end.
Ericsson, Qualcomm and Singapore operator SingTel recently demonstrated peak speeds of 260Mbps using FDD/TDD carrier aggregation. The cellcos’ managing director for networks, Tay Yeow Lian, said: “We are already using dual-layers of LTE FDD with LTE-Advanced carrier aggregation, and we know the performance of our mobile network is critical to ensuring that our customers have a great experience on their smartphones and tablets. Therefore, we will continue to explore the future option of bundling even TDD spectrum into the LTE carrier aggregation to expand network capacity, allowing our customers to enjoy consistent high data speed download even at high traffic areas.”
The technical demonstration took place at the SingTel office in Singapore on October 3, using a Qualcomm Snapdragon 810 processor and Ericsson RBS6000 base stations, aggregating 20MHz of FDD and 20MHz of TDD spectrum.