Thursday 31 October 2013

Small Cells and Wi-Fi in 3GPP Releases

Here is a quick chart of features of Small Cells and Wi-Fi introduced by 3GPP in different releases. Please feel free to suggest anything that has been missed out.

Monday 28 October 2013

Optimizing Small Cells and the HetNets

Came across a whitepaper from JDSU, not so new but its got some interesting stuff. In an earlier post here, we saw the challenges for small cells deployment, the picture above shows another view.

Another interesting item is adding intelligence to small cells. This can mainly help by reducing the traffic back to the core. This also relies on predicting the user behaviour which could be a challenge in it self. There is a lot more interesting stuff in the paper, which is embedded below:

Sunday 20 October 2013

New opportunities in Carrier Wi-Fi & Wi-Fi offload

Interesting webinar from Maravedis-Rethink exploring the new opportunities with Carrier-WiFi and Offloading. Embedded below is the video and presentation:

Monday 14 October 2013

The right technology for different Enterprises

The enterprise deployments seem to be hotting up. Back in May, Telefonica O2, Germany, announced 'Signal Box', their enterprise femtocell based on UMTS to improve indoor coverage. Recently Vodafone in Netherlands announced that they are offering enterprise small cells from Spidercloud. There was an interesting article in Fierce Wireless about the enterprise small cells opportunity. The relevant part is reproduced below:

So it is perhaps no surprise that the industry has turned its attention to the enterprise segment, one which has less price-sensitive customers with higher demands for coverage, capacity and sophisticated mobile services. Enterprise small cells are much more in evidence, such as ip.access' nanoCell, which can be deployed in conjunction with Distributed Antenna Systems (DAS) in enterprise environments, and many vendors now focus on how to position small cells to penetrate this high-value segment. This is the new hot segment, opening up a rich service portfolio to offer to enterprise customers, a market that has been a tough nut to crack for mobile operators.
Combining enterprise small cells with network-based management applications potentially enables operators to add significant value to this segment by bringing new mobile services and features. These could include dedicated voice capacity, mobile unified communications, mobile call recording, local switching of voice traffic and context aware services, as well as the fast-developing field of network analytics.
However, there are also several key enterprise challenges that are significantly more complex than those found when offering femtocell services to consumers, such as managing inter-cell interactions, delivering consistent coverage over larger areas than a home or SOHO, as well as typically needing to support increased user numbers with higher user mobility. All of which means that as small cells become part of mobile operator strategies, there is a rising need for common guidelines and best practice, for products and deployment and Release Two provides an "all you need to know" guide to help operators deploy enterprise small cells. This follows on from Release One, a similar exercise centered on consumer femtocells, unveiled in February at Mobile World Congress. Release Three focuses on urban small cells and is scheduled for release at Mobile World Congress 2014.
Infonetics believes that deploying small cells within the enterprise segment represents a huge opportunity for mobile operators for several reasons:

  • the mobile operator has the opportunity to increase indoor coverage within an enterprise campus, both indoors and outdoors;
  • enterprise IT architects are seeking to move all personal communications services on to mobile devices and "unwire" their organizations, particularly as employers increasingly have BYOD policies;
  • as they do so, there is a need to "mobilize" some of their existing communication services such as IP PBX, Centrex, IP VPN etc.; this also takes away the emphasis on "buying small cells" and places it on "improving/mobilizing communications" as the cost of those cells can be bundled with those services
  • in deploying infrastructure within the enterprise premises, the operator has the chance to place "golden handcuffs" on the enterprise and pull them into an extended length of contract; one might argue that BYOD means that there is a need to support multiple mobile operators not just one selected by the enterprise, but in reality if one operator offers to provide small cell coverage, there is usually a positive migration to that network by employees (even if it is a case of BYO SIM);
  • IT budgets are shrinking and there is increasing pressure on enterprises to outsource communications as they don't have the staff to deal with the complexity and rate of change, so there is the potential for operators to take on network operations for their high value customers
However, to date the enterprise has largely proven hard to target effectively. Traditionally this market segment has been underserved, as options such as DAS have only been available to larger scale enterprises due to the considerable costs involved. There are specific challenges of using DAS, which require a room in which to deploy a macro/pico BTS, and then the installation of industrial-grade coax to pipe the signal around the building, and sometimes requires an upgrade to the building's power. This is often invasive, expensive and time-consuming.

Recently I came across an interesting report on Enterprise small cells from Maravedis-Rethink which is embedded below:

Tuesday 8 October 2013

Super Macros and HetNets

The other day I read the following on Light Reading:

UK operator EE wants to turn its existing macro cell sites into "super macros," according to Andy Sutton, the carrier's principal network architect, speaking at the recent Base Station conference in London.

EE 's plan to super size its macro cell sites fits in to a broader Heterogeneous Network (HetNet) strategy for adding capacity and extending coverage. The operator rolled out the first LTE network in the UK last year, and has now covered 55 percent of the UK population and has 1 million 4G customers.

"Super macro is the first step toward building a HetNet,” said Sutton. “Evolving the macro is the most cost-optimized way to adding capacity into our networks."

But what makes a macro super?

According to Sutton, a super macro would typically have multiple radio access technologies (RAT), three-to-six base station sectors, and operate in multiple frequency bands using carrier aggregation techniques. It could be a standalone base station or a hub for subtended, smaller micro cells. He added that infrastructure sharing is vital to the strategy as well.

Once the operator has sufficiently beefed up its macro cell sites, then it can look to smaller cells to be deployed indoors and outdoors in hotspots or cell edge locations. Sutton described a small cell deployment as an "underlay" to the super macro.

The term super macro isn't exactly new, but the fact that operators are talking about it now indicates just how much more they are looking to do with their existing radio access network (RAN) infrastructure before introducing new small cells or while planning a small cell deployment.

"Within the super macro concept, there's quite a lot operators can do to improve performance," says Heavy Reading senior analyst Gabriel Brown.

Along with adding sectors, using more spectrum bands, or employing carrier aggregation, Brown also includes in the super macro concept using 4x4 and 8x8 MIMO, active antenna systems, vertical sectorization, or beamforming.

The advantage of improving macro sites is that many of the basic elements that go into the total cost of ownership of a cell site are already in place, such as power, real estate rental, and backhaul, according to Brown.

I remember the folks from Ericsson mentioning about Super macros but I had not given any thoughts to it. Well, I went back to see what they have been talking about and found this:

Since it was not very clear, I found some additional information from an NGMN presentation as follows:

In order to reduce UEs’ frequently handing over between neighboring cells when moving at high speed which results in voice service or data download breakouts in the railway/subway/highway scenario, different RUs in different sites using C-RAN architecture can cooperate with each other and many macro chain cells can be combined to a super macro cell. In this burst communication scenario, network performance has higher priority than network capacity.

So my suspicion is that Super macros would be ideally using C-RAN where it would be possible to combine what could be many macros into a 'super macro'.

If I start thinking about it, there can be additional uses of the super-macro:

  1. Carrier aggregation - scenario 4 - see here. In Release-10 scenario 4 is not possible because of different timing advance requirements but this has been resolved in Rel-11. Super Macros can be useful in this scenario where small cells provide capacity, super macro provides coverage and reduces the need for constant handovers, etc.
  2. NCT - related to the above - see here - again, smaller cells (metrocells/microcells) using NCT to provide capacity, super macro for coverage
  3. Phantom Cell - see here - related to 1 and 2 above, super macro is the coverage, connectivity and mobility layer, small(er) cells are phantom cells that provide higher data rates. 

Anyway, a lot of information is pure speculation so feel free to add more info or correct my understanding.

Thursday 3 October 2013

Ericsson's Radio Dot System

Just come back from the SON conference, I noticed that Ericsson's dot system was discussed heavily on what it is and how its different from Alcatel-Lucent Light Radio and NSN's Liquid radio. Here is a bit of detail from the Ericsson press release:

Ericsson Radio Dot System is compact and offers flexible mounting. The device weighs 300 grams, is the result of two years of research and development, and incorporates 14 patents. It introduces a revolutionary antenna element, or "dot," which delivers mobile broadband access to users. Because of its convenient size, scalability, and compelling evolution path, this product caters to different kinds of users in medium to large indoor locations, and aims to address operators' needs of offering a complete indoor solution.
Dots are connected and powered via standard internet LAN cables (Category 5/6/7) to indoor radio units that link to a base station. Radio Dot System leverages the same industry-leading features found in Ericsson's macro base station. Deployments and upgrades are simple, addressing growing capacity and coverage requirements. Thus the users' experience is consistent wherever they go and the indoor network evolves in lockstep with the outdoor network. Ericsson Radio Dot System supports integration with Ericsson's carrier Wi-Fi portfolio enabling features such as real-time traffic steering to ensure the best user experience across both Wi-Fi and 3GPP networks. 

There was a bit more detail available from a discussion in the Linkedin HetNet group here. My edited version of the discussion:

A related patent by E///, on how they may be transporting digital I&Q (essentially CPRI) between RadioDot and IRU by introducing APL.

<The layers, from top to bottom, are CPRI layer, Adaptive Physical Layer (APL) and the Ethernet Physical (PHY) layer....

APL is a newly introduced layer for adapting and packetizing the data from CPRI layer to comply with the 10GbE PHY data frame format... >

Based on my understanding, this is a standard Macro Cell with all functionality available from the day it is launched except that the PA+LNA+Filters+Antenna Panel (DOT) is extended remotely over Cat V cables. You can have multiple dots connected to a single cell based on the requirement as per DAS architecture. Now the Base Band Board and Radio Head may be centralized for the building. So this architecture may not really comply with C-RAN concept, so lots of features of C-RAN can't be supported with this. This is the very first step towards a centralized LTE DAS architecture for a big network. I think this is going to be a costly solution because of the architecture. One thing which is not known if this is based on SoC or the standard DSP FPGA based solution, because cost will be drastically low if it is based on SoC. Now it is interesting to see how the big carriers are taking this, since it is going to be launched in the second half od 2014, probably a very flexible hybrid DAS solution based on SOC will be available from many of the leading Small Cell Solution providers.

After my discussion with Ericsson executive, he explained the dots can be configured as different cells than the macro cell (in case the BBU is used for macro as well as indoor cells as posted in link to Ericsson document on page 11). 

I am very skeptical of this solution. It's a very nice marketing initiative - but I will be surprised if it achieves any serious market share. Mobile operators are mostly not capable of building the sort of networks that will require the dot - so deploying such a thing requires a change in mindset that is not easy to achieve. Secondly, the "dot" does nothing beyond what a $300 802.11n or 802.11ac Wi-Fi carrier class access point can deliver. And I'm sure $300 is not the Ericsson price for the dot. So I think the dot will find a very small market indeed. The much more interesting market is Wi-Fi offload at a fraction of the price. :-)

Embedded below is a presentation explaining the dot system: