Operators are casting around for almost any source of additional LTE spectrum, even in bands previously regarded as unsuited to mobile communications. New techniques such as carrier aggregation and supplemental downlink will help to make unpromising frequencies usable, but with every addition of new mobile services to the spectrum, the more the potential for unforeseen interference problems.
The threat of interference is real in many adjacent bands, but can also be exaggerated for political effect by companies seeking to preserve an incumbent position, so the debates tend to be resolved far more slowly than the growth in mobile data demand would merit. The most famous recent example of the new capacity versus interference dilemma was, of course, seen in the LightSquared fiasco in the US, when the eagerness to harness an underused band ‘ mobile satellite L-band spectrum ‘ for LTE apparently overrode warnings about proximity to existing GPS services. Or so argued the GPS incumbents, whose objections finally drove the FCC to refuse LightSquared permanent permission to pursue its plans.
That does not mean that mobile satellite spectrum is off the agenda as a promising addition to LTE resources. Some of its frequencies lie further from GPS ‘ Dish Network is considered likely to win FCC permission to run LTE in its S-band holdings. And in other parts of the world, the L-band itself is still being eyed as a 4G contender. Last month, Orange France won approval from its regulator, Arcep, to test L-band spectrum for LTE. It will use the supplemental downlink (SDL) mode, also being adopted by AT&T in order to use the unpaired 700MHz frequencies it acquired from Qualcomm, to add capacity to its main paired 700MHz LTE airwaves.
SDL was standardized for HSPA in 3GPP Release 9 though is likely to be used mainly in LTE. It allows an SDL carrier to operate in a different band from the main cell, and in unpaired frequencies. In LTE, SDL operation is extended by bonding up to three carriers in the unpaired band with the serving carrier(s) in the paired spectrum.
Orange, working with Ericsson and Qualcomm, will test the technology in the city of Toulouse. They will evaluate the potential to use additional frequencies to boost capacity of UMTS and LTE networks, in this case with L-band spectrum in the 1452-1492MHz band. The trial will be carried out during the first quarter of 2013 and will harness Ericsson base stations and devices running on Qualcomm chipsets ‘ all of them using the supplementary unpaired frequencies in the L-band for downlink operations, combined with a traditional paired carrier at 2.1GHz. The roll-out of the trial network will start towards the end of this year.
Subject to interference concerns, CEPT (European Conference of Postal and Telecommunications Administrations), the agency responsible for the harmonized use of the L-band in Europe, has made positive noises about refarming some of the spectrum ‘ currently mainly used for digital audio broadcasting – for mobile broadband. Its eventual recommendations will feed into a similar evaluation at global level, by the ITU-R, which will present its findings at the next World Radio Conference (WRC) in 2015.
Last year, Plum Consulting director Phillipa Marks called on regulators worldwide to refarm spectrum in the 1.4GHz band. ‘There is a whole patchwork of legacy applications [in the 1.4GHz band] that have been mostly sitting idle for the last 30 to 40 years,’ she said. ‘The refarming activity for 1.4GHz is minor compared to almost any other frequency band,’ she added, indicating availability in many areas of Europe, Asia and Africa. She proposed that the 40MHz of spectrum which could be freed up in this band should be used to provide a supplemental downlink to accelerate download speed.
The UK has already auctioned some 1.4GHz spectrum for mobile use, back in 2008, when it was snapped up by Qualcomm for £8.33m. At the time, that fitted in with the chip giant’s MediaFLO plans for unpaired frequencies, but it is not now clear what the company will do with its assets. However, its European chief, Enrico Salvatore, was pushing UK regulator Ofcom only last week, urging it to get LTE rolled out quickly. So far, the LTE auction has been delayed by carrier disputes and fears over interference with TV signals. Salvatore’s vocal comments suggest that the US firm may be keen to see 4G roll-out start, so it can impress upon the cellcos the attractions of its L-band holdings for supplemental downlink.
It is not just in satellite spectrum that techniques like refarming and SDL can spark stand-offs with other carriers around interference fears. AT&T’s complicated plan to combine different frequencies within the 700MHz band in the US has added a new layer of controversy to already heated disputes over this spectrum and its fragmented nature. Already, small carriers are lobbying the FCC to force Verizon and AT&T to make their LTE devices, which run in different parts of 700MHz, work across the whole band. Ownership of this spectrum is very fragmented. After auctions held in 2003 and 2008, Verizon gained Block C (mainly in 747MHz-787MHz), the only national licence, but one that carried open access mandates with its 22MHz of capacity. No fewer than 100 other companies bought licences in the A, B and E Blocks, mainly in band class 12 (Block D, earmarked for public safety, went unsold). AT&T was the other big buyer, with 150 licenses in the B Block (704MHz-746MHz or band class 17), which generally covers metropolitan areas. AT&T also acquired Qualcomm’s spectrum, which had been acquired in 2003 for the now-defunct FLO TV service in the Lower D and E Blocks.
To make the issues around roaming worse, AT&T is also trying to segregate its network by creating its own band class surrounding its B and C Block licences ‘ which it says is necessary to protect the Qualcomm frequencies from interference from the A Block and therefore make SDL viable to augment capacity. That prevents it roaming with small carriers in the A Block, though its stance will become even more complicated if it buys the licences Verizon plans to sell off in the A Block (should Verizon’s own purchase of AWS spectrum from four cablecos be approved).
Some of the most detailed arguments against mandating a device which would work across the whole 700MHz band ‘ on the grounds of interference risk ‘ have come from Qualcomm. Of course the firm has an interest in promoting its own solutions ‘ it promises a seven-band solution in future ‘ but it also has considerable experience from its MediaFLO adventure. It warned the FCC that issues with Channel 51 and the unpaired D and E Blocks would all need to be resolved before unified 700MHz devices would deliver acceptable performance.
When it was deploying MediaFLO in its D and E Block, Qualcomm said it discovered that actively used E Block stations would cause such high levels of interference that B and C Block receivers would not be able to get signals ‘ and that the best current solution is to use the A Block spectrum as a guard band, as AT&T is doing. The regional carriers’ band class 12 enables A block spectrum for active use, and does not provide any protection against E Block interference, argued Qualcomm. In addition, where Channel 51 is active, band class 12 A block deployment is forbidden by the FCC.
However, one of the most vocal of the regional carriers, US Cellular, is putting its money where its mouth is on interference. It is acquiring various licences in the Lower A Block from fixed line carrier CenturyLink, covering selected markets in Wisconsin, Missouri, Illinois, Nebraska and Oregon. As reported by RCRWireless, the deals indicate that some operators are becoming less preoccupied with ongoing issues in this part of the 700MHz band, and are choosing to accelerate their roll-outs even before the roaming and interference disputes are concluded rather than fall behind in mobile broadband.
US Cellular was part of a group of regional cellcos which recently claimed that, despite AT&T’s and Qualcomm’s statements, there were no significant interference issues between the lower 700MHz bands and TV broadcast channel 51. Some small operators imply AT&T is just trying to avoid roaming with rural cellcos, and US Cellular has already used its A Block holdings to deploy 4G in some markets (it has more than 150 A and B Block licences via its partnership with King Street Wireless).
Verizon has offered to offload its holdings in A and B Blocks if its bid to acquire AWS spectrum from four cablecos is permitted by the FCC. US Cellular’s recent actions put it firmly in the frame to acquire those Verizon assets, at least where they fall within its operating regions.
The FCC opened a discussion on 700MHz roaming and interference in May, and has been bombarded with petitions from smaller carriers to force the big two players to support roaming with all 700MHz bands. Interference is not the only argument which has been deployed against the small cellco. Verizon has also pointed out the cost and complexity of incorporating so many radios in affordable devices ‘ potentially a dozen or more bands would need to be supported to enable its own multiband roll-outs, international roaming as well as domestic interoperability.
As epitomized by the 700MHz debates in the US, refarming and supplemental downlink will be vital to achieving the spectrum capacity needed for mobile broadband, but these approaches will only heighten the interference debates around the world, not to mention the challenges for chip and device makers, with an ever growing total of frequencies to support. This will be the price to pay for accessing sufficient spectrum to meet the world’s wireless data demands.