Small Cells, Heavy Lifting

5:45 PM -- You might notice that vendors like Alcatel-Lucent (NYSE: ALU), Ericsson AB (Nasdaq: ERIC), Nokia Networks and others are all pushing the virtues of a distributed architecture, often using small cells to increase the performance of 4G Long Term Evolution (LTE) networks.

The reason for this is simple: The closer you can get the radio to the end user, the faster the uploads and downloads. This is a particular consideration in high-traffic city areas where there are lots of users texting, emailing and watching video.

There is, however, a sting in the tail for carriers thinking of deploying these networks. Just as it seemed they were getting a handle on deploying enough capacity to backhaul faster 3G and nascent 4G networks, along came small cells to throw a spanner in the works again.

They present a distinct set of challenges:

    Backhaul connectivity: It is not often going to be practical to run fiber up a lamppost to a small cell base station. Microwave and -- possibly -- free space optics could prove to be an alternative.

    Cost: Clearly, carriers won't want to spend $10,000 to backhaul a $1,000 radio.

    Capacity: Already the industry is looking at clusters of radios that each carry hundreds of megabits of data traffic simultaneously. Small cell backhaul will still need to be a fat pipe.

    Ease of deployment: Operators may need to add small cells fast to keep up with user traffic. The backhaul will need to be right there behind it.

There are a number of companies already trying to deal with these issues. Still, I think it's worthwhile to remember that the next time you hear a vendor wax lyrical about small cells, there's still a lot of heavy lifting that has to go on behind the scenes.

— Dan Jones, Site Editor, Light Reading Mobile

krishanguru143 12/5/2012 | 4:53:06 PM
re: Small Cells, Heavy Lifting


krishanguru143 12/5/2012 | 4:53:05 PM
re: Small Cells, Heavy Lifting

Nope, I had caps on some letters as well, so what did your server do to my post is the bigger question.

ockham 12/5/2012 | 4:53:05 PM
re: Small Cells, Heavy Lifting

ironic remark but indeed pop-out a key success factor for SC (Small Cells) which is the availability of cheap MBH.

Why you think a MBH should cost 10K if SP can sell (making money) a 40 Mbps internet connectio at 30-40 USD/month?

SC success will required a similar approach, BH through FTTH/DSL at &lt;10 USD/cost/month.

Fully feasible IMHO.

joset01 12/5/2012 | 4:53:05 PM
re: Small Cells, Heavy Lifting

Typing on the iPhone again, Ian?

psenior 12/5/2012 | 4:53:03 PM
re: Small Cells, Heavy Lifting

Agree with all of your points. TDD backhaul is best as DL/UL symmetry can be chosen to match traffic split.

Picos cells are almost always in NLOS conditions, so OFDMA is the right backhaul radio technology to make it work with a lot of multi-path. This helps MIMO anyway.

My understanding is that business case for these integrated Backhaul + Access outdoor Picocells works when the whole thing costs less than $6k USD.


allip 12/5/2012 | 4:53:03 PM
re: Small Cells, Heavy Lifting

Good Points Paul...

Few additional items which could become a requirement / diffrentiator besides the cost and power

1. Some Carrier Ethernet based OAM

2. Sync solution, given the fact the most sub 6Ghz soln would be TDD

3. Ethernet switching with some guranteed QoS for the controller node in P2MP config

4. Would be interesting to see if FSO can meet the pricing of the lost cost microwave solutions either sub 6Gh or greater with the ASP of a P2P all outdoor solution available for uder 4K soon...

There seems decent activity in the small cell backhaul space with starups like Blinq, Taqua, Tarana Wireless and the ofcourse the tradiontal uwave backhaul folks also working on it and interestingly all most likely using the same modem vendor, with similar architecture.


psenior 12/5/2012 | 4:53:03 PM
re: Small Cells, Heavy Lifting


Couple of other things for your list.

Wireless Backhaul is ideal, but needs to work non-line-of-sight, that rules out 60 GHz stuff. The best band is typically 3.5 GHz (lots of cheap spectrum available). Backhaul should be at least 10bits/Hz/sec.

Backhaul need to deliver end-to-end QoS. This is critical for all LTE small cells. This is easy with P-P links with very high capacity, but non-trivial for P-MP backhaul that is contended.

The "Pico" base stations need to be small, but still be powerful. Everyone things that 2x1W is the optimal power.



OpEd 12/5/2012 | 4:52:59 PM
re: Small Cells, Heavy Lifting

Beyond backhaul, I heard operators asking at the LTE Asia 2011 conference a few weeks ago about the deployment costs of small cells, given how far (or little) SON has been implemented.

Another topic of interest was securing communications of pole-top and other small cell installations.&nbsp; 3GPP recommends IPsec for bearar and control plane traffic in cross (organization) domain communications.&nbsp; Moreover, the combination of the All-IP network and physical access to small cell baseband units provides additional impetus for securing communications with the core and adjacent radios.&nbsp; IPsec is an eNodeB feature requirement, so this should not add costs to the radio, but an overall security solution should be considered.



thephoneman 12/5/2012 | 4:52:56 PM
re: Small Cells, Heavy Lifting

Just spit-balling, but there are already millions of deployed (and connected) "small cell" sites. &nbsp;Its called WiFi. &nbsp;I realize this isn't the answer for truly mobile applications but, as the research I've seen says, most traffic isn't truly mobile. &nbsp;Meaning people aren't watching videos while driving but waiting until they get somewhere to do so.

If that's the case, wouldn't it make more sense to somehow standardize the configuration of, authentication by, and (more importantly) the billing from existing WiFI access points?

anyfinetworks 12/5/2012 | 4:52:36 PM
re: Small Cells, Heavy Lifting


I like your thinking. :) In fact we've been working an a system to accomplish what you're suggesting for the last few years. Only difference is we use existing standards and instead run the Wi-Fi protocol (802.11) over an IP tunnel to the operator core (or your own home gateway) in order to ensure perfect security and a completely transparent user experience.

And when I say transparent I really mean it; you can e.g. have the exact same user experience as you do at home (WPA-PSK encrypted home Wi-Fi) when mobile. There's no registration process, no manual sign-on, no software to install on the device, no settings to change, nothing. It just works.

If you're interested have a look at http://www.anyfinetworks.com/about/solutions.

Best regards,


Sign In