Sprint's Magic Box

Is Sprint doing Small Cells? That's a question probably asked too many times. Back in January, their COO Günther Ottendorfer said the company’s small cell partners conducted a range of trials last year in order to determine fast and efficient methods to deploy small cells, a situation he said led to some misunderstandings in the market. However, he said those trials are largely behind the carrier and that he expects the carrier’s small cell efforts to expand this year.

“There was a learning process in 2016. We did a lot of trials in the beginning. We had some trials that led to misunderstandings, when you have a lot of boxes there because you were trialing different things, different—for example—transmission methods,” said Ottendorfer, Sprint’s chief operating officer for Technology, in a recent interview with FierceWireless. “But now we have streamlined the concepts and so I’m very confident that with streamlined and very elegant small cell solutions we will have a good rollout this year.”

They again mentioned about their small cells commitment at MWC. Finally this week, they announced the Magic Box.

Sprint has billed it as "World’s First All-Wireless Small Cell". This is a point where I would disagree with them, mainly for two reasons.The first being that for an all-wireless claim, they have to get wireless power to the small cell and secondly, this has already been done for a while. I have explained about In-band backhaul here and have provided examples of how Parallel Wireless has been using this for a while.

The Magic Box is made by Airspan and is 4G/LTE only in band 41 (2500 MHz TD-LTE). One of these units provide an average coverage of 30,000 square feet indoors and can benefit adjacent Sprint customers inside the building. The signal can also extend coverage 100 meters outside a building, benefiting customers in nearby buildings and improving street–level network performance. It does not use the closed subscriber group (CSG) feature hence anyone can camp on it and use it.

Source: FierceWireless

Sprint has a large amount of 2.5GHz spectrum available, as a result they are able to use dedicated spectrum for the Magic Box. This ensures that interference is kept to minimum. They also announced the availability of HPUE that will allow this band reach to improve. See my blog post here for details.

“It’s a far cry from just a repeater,” he said, explaining that it improves the efficiency of the network as long as it has a good connection to the macro cell. It will work with any Sprint phones using 2.5 GHz. The backhaul channel uses 2.5 GHz or 1.9 GHz, but ideally it would use 2.5 GHz because that offers a lot more capacity.

The Magic Box includes self-organizing network (SON) capabilities and operates on its own channel in Sprint’s spectrum, allowing it to decrease the noise level and increase the capacity of the overall system, which is the big difference from repeaters, explained Sprint Technology COO Guenther Ottendorfer.

Some of the details I couldn't find but hopefully some of the readers would know and can answer are:

• Whats the power output of these small cells?
• I am assuming they will support VoLTE calling for voice - even though generally that feature is transparent to small cells?
• Does the small cell radiate a single 20MHz channel?
• Does the backhaul do carrier aggregation?

Further Reading:

• Sprint Fiscal 4Q16 Results - some good insights on small cells
• CTO Blog: The World's First All-Wireless Small Cell - Sprint Magic Box
• Sprint Announces World’s First All-Wireless Small Cell – The Sprint Magic Box
• Can Sprint’s New “Magic Box” Live Up to Its Name? - Wireless Week
• Sprint introduces Magic Box, and it’s not a simple repeater - Fierce Wireless
• Sprint's new MagicBox is an LTE-A UE Relay node, made by manufacturer Airspan.

Unlicensed LTE (LTE-U) for Small Cells

I first wrote about LTE-U (or U-LTE as referred by others) back in December last year pointing towards the Qualcomm presentation here. As things move fast in our industry, quite a few things have happened in the last 6 months. Huawei did a demo of U-LTE in Mobile World Congress this year and LTE-U has been a constant topic of discussion in various 3GPP meetings. A half-day workshop is due to take place in June to discuss this topic further. In the meantime I have summarised some of the discussions that took place (unofficially?) in Jan 2014 between the interested parties.

To be clear, the discussions for LTE-U are centred on the 5GHz unlicensed spectrum. As you can see from the picture above, there is a massive amount of spectrum that is available, either free to use (unlicensed) or in a lightly licensed way.

There are strict rules and regulations in place to make sure this license is not misused or monopolised. There will be a need for Dynamic Spectrum Access (DSA) techniques that I have discussed here. The current LTE standards do not have a DSA inbuilt and hence referred to as "Rude". The following is from a recent Light Reading article.

The technical concern with LTE-U, as Peters describes it, is that LTE is a "rude" technology. WiFi includes a "politeness protocol" that LTE lacks, meaning that WiFi will back off if it senses interference from other users. Eventually rude ol' LTE operating in WiFi's polite bands could take over the band.

The 3GPP called another unofficial meeting in January to discuss concerns around LTE-U, which also included the potential effect on the value of licensed spectrum, the need for international harmonization of the unlicensed bands used for LTE-U, and whether the technology would be for downlink only or uplink as well. The group met again in March, primarily to work out timing for the new technology's deployment.

Huawei presentation explains why unlicensed carrier offloading, see the embedded presentation for details.

NTT Docomo shows the various deployment scenarios and also lists the regulatory aspects, especially in Japan. See the complete presentation below:

Nokia has even gone ahead and done simulations for different scenarios. The scenario above shows LTE deployment in the same unlicensed band as WLAN as you can see, the results are similar to the interference in WLAN-WLAN case.

There is also a roadmap to how LTE-U standardisation will work in 3GPP, hopefully after the workshop in June, we will probably hear more.

As expected, some of the operators with heavy investments in WiFi (like AT&T) have some reservations on LTE-U. Some analysts on the other hand are sceptical on how much savings there would be, taking the interference into account. Note that spectrum is just one part of deployment costs, there are many other factors to consider. Personally, I don't have an issue whether this will work or not, it definitely would do but with all the advancements in LTE-Wi-Fi Interworking, I think we may be able to do a better job with just selective deployment of LTE-U and using technologies like MAPCON, IFOM, Hotspot 2.0, etc.

Added on 9th July 2014

3GPP held their workshop on unlicensed LTE on 13/06/2014. See the news on 3GPP website here. All documents are available here.

Dynamic Spectrum Access (DSA) techniques for Small Cells and Wi-Fi

Picture Source: Analysis Mason

There is a lot of spectrum which is used sparingly or is kept reserved for unlicensed or shared access. Any party that wishes and is allowed to use this spectrum has to co-ordinate with the license holders or others in similar situations. Hence we have different access mechanisms which are collectively called as Dynamic Spectrum Access (DSA) techniques.

An article by Analysis Mason on this topic suggests the following:

The term DSA has come to encompass a number of different approaches and techniques that aim to increase the utilisation of the radio frequency spectrum. At its most ambitious, it is hypothesised that cognitive and software-defined radios could intelligently choose when to transmit, so as to avoid other radio transmissions and also to avoid causing undue interference to fellow frequency users. Short-term propositions include near-real-time spectrum assignment in certain bands and greater use of long-term secondary spectrum leasing to authorised spectrum partners.

DSA is therefore all about making better use of radio spectrum through re-use of 'idle' bandwidth, being either frequencies that are not used in all locations, in which other systems could be deployed, or frequencies that are only used intermittently, and which could therefore be re-used outside these times. These 'gaps' in utilisation, which provide opportunities for DSA, arise for a number of reasons.

• Coverage: a licence holder might not be using its allotted licence in a specific region.
• Time: an area of spectrum might by less-frequently required at different times during a day (or on longer timescales).
• Lack of service users: there may be a limited number of subscribers taking advantage of a service.
• Licence technical parameters: the regulator may have mandated that a piece of spectrum can only be used for a specific purpose, while other technologies emerge during the life of a licence that can use the same spectrum.
• Pragmatic under-utilisation to prevent interference: empty guard bands are placed between spectrum bands to stop transmission leakage to prevent interference, which could be re-used by systems that have the appropriate characteristics to avoid interference.

One of the overarching drivers for DSA is to help overcome spectrum shortages – particularly noting that under-utilised bands may exist across a relatively wide range of the spectrum. Even in the economies where wireless communications have developed the most and usage restrictions have been removed, thus making spectrum use as flexible as possible, spectrum under-utilisation is still considered to be widespread.

The Cisco vision on the other hand seems far too optimistic and suggests the following:

TV White Spaces (TVWS) are spectrum allocated to TV broadcasts, but not being used in a given geographic location. TVWS radios allow for use of white space spectrum for unlicensed wireless access.

Authorized Shared Access (ASA) or Licensed Shared Access (LSA) allow a secondary licensee to use the “shared” spectrum when the primary licensee is not using it.

The United States Federal Communications Commission (FCC) has proposed a three-tier model for shared access in the 3.5-GHz band. Tier 1 would be for incumbent federal agencies, including military radar users. Tier 2 would be authorized prioritized access similar to ASA and LSA. Tier 3 would be generalized authorized access, which is similar to unlicensed access.

A number of DSA technologies already exist or are in exploration.

Geo-location, database-based spectrum sharing techniques have the most traction as a practical approach to spectrum sharing. Devices that want to use shared spectrum must geo-locate themselves and consult a database to determine what spectrum is available.

The geo-location database manages the spectrum resource allocation based on predefined policies and availability to ensure the primary licensee is not impacted. An enhanced version of the geo-location database system—called a Spectrum Access System (SAS)—is the basis for the FCC spectrum-sharing proposal in the 3.5-GHz band.

A second technology is cognitive radio, which senses and monitors the radio environment. This includes knowing the location and policies for self-regulation. Dynamic Frequency Selection (DFS) and Transmit Power Control (TPC) are cognitive radio techniques that allow co-existence with radar and satellite systems.

Another technology is Software-Defined Radio (SDR), which allows devices to adapt to local radio conditions and use the appropriate radio frequencies.

I came across this very interesting whitepaper by MIT that details all the DSA techniques and its progress. Paper embedded below:

New Models for Protected Shared Access - Toward More Efficient Spectrum Management from Zahid Ghadialy

We will discuss in the future post how the DSA techniques could be useful for using Small Cells in the unlicensed spectrum (a.k.a. LTE-U).

LTE-U: First step towards 'Operator-neutral' Small Cells

3GPP member companies have decided to take the Small Cells vs WiFi fight back into the Wi-Fi camp by unveiling the initiative to have LTE in the unlicensed band (officially known as LTE-U but another interesting term uLTEA has been proposed unofficially too). 

Mike Roberts wrote an interesting post on ITM Blog:

Qualcomm has launched a radical initiative to deploy LTE Advanced in the unlicensed 5GHz band, which has traditionally been the preserve of Wi-Fi devices. If the initiative gains support, it could be a hammer blow to carrier Wi-Fi, though Qualcomm says it is still a strong supporter of all types of Wi-Fi, given its Qualcomm Atheros division, which is a leading producer of Wi-Fi chipsets. 

Qualcomm CEO Paul Jacobs unveiled the company’s initiative for LTE Advanced in unlicensed spectrum at its financial-analyst day in New York on Nov. 20. It was interesting timing, considering that the main annual carrier Wi-Fi event, the Wireless Broadband Alliance’s Wi-Fi Global Congress, was taking place on the same day on the other side of the world, in Beijing. 

The chipset giant followed the announcement by its CEO with detailed presentations on the initiative at Informa’s LTE North America event in Dallas on Nov. 21-22. The essence of Qualcomm’s proposal is that it is complicated to integrate LTE and Wi-Fi to create carrier Wi-Fi, so a better option could be to use LTE Advanced to extend LTE into unlicensed spectrum, thus eliminating the need for Wi-Fi in that context....So how does Qualcomm plan to square this circle and offer reliable mobile services in unreliable unlicensed spectrum? By using LTE Advanced to create a hybrid system operating in both licensed and unlicensed spectrum at the same time. Specifically, Qualcomm is proposing a system that uses the carrier-aggregation (CA) feature of LTE Advanced to aggregate licensed LTE spectrum with unlicensed spectrum in the 5GHz band. The unlicensed spectrum, which would be aggregated with licensed spectrum on the downlink only, would only be used for data services. The licensed spectrum would be used for uplink and downlink and would support network control as well as voice and data services. Given the power limits placed on devices using the unlicensed 5GHz band, uLTEA would be used mainly in small cells, similar to Wi-Fi.

'All about 4G' did a good quick write-up on this topic from standards perspective, as follows:

Qualcomm has recently floated the idea of deploying LTE in unlicensed bands, particularly focusing on the 5GHz band, which is currently used mostly for WiFi. According to a document (RP-131635) submitted to the upcoming 3GPP plenary meeting, the proposal is to deploy LTE as Supplemental Downlink (SDL) in 5725-5850 MHz in USA, with the PCell (Primary Cell) always operating on a carrier in a licensed band. Verizon has also submitted a Work Item Proposal (RP-131680) to to introduce the new band for SDL usage. There’s also a Study Item proposal from Ericsson (RP-131788) is the rapporteur to study the modifications necessary to the LTE radio.

One of the big issue (though small cell vendors often play it down) that has been delaying the rollout of Small Cells has been interference management in co-channel deployment cases. The operators are worried about ad-hoc small cells creating interference with the properly planned and optimized Macro cells (though in practice its not as bad as they think it is). With LTE-U, this issue can be put to rest.

I can see LTE-U, if adopted, can also help create new business models around Small cells. Neutral Small cells, deployed in the hotspots like stadiums, shopping malls, etc. can serve users of all networks. The signalling can take care of each user connecting to its own network using backhaul.

Femto as a Service (FaaS) and Small Cells as a Service (SCaaS) would probably become a common approach as there would be no concerns about which spectrum is being used. 

There would be a small concern about interference between Wi-Fi and Small cells. Most of the existing Wi-Fi and other devices that use unlicensed spectrum use the 2.4GHz band with a few using 5GHz band now. If LTE-U uses the 'Listen Before Talk' approach, other technologies can be enhanced to probably do the same. Another thing worth remembering is that there is nearly 300MHz of spectrum in 5GHz band available universally. In certain countries the available spectrum can be as much as 500MHz. This should be enough for LTE-U as well as other technologies in the unlicensed spectrum.

The complete Qualcomm presentation is embedded below and is available to download from here: