Our Director of Technology, Simon Saunders, joined a University of Oxford Forum discussion to respond to some frequently – and not so frequently – asked questions regarding small cells.
The transcript follows for your interest (this has been gently edited to align the threads of the discussion. The original is at http://linkd.in/11S10VW)
ForumOxford Free Friday chat – 10 May – Dr Simon Saunders joins us to discuss ‘Small Cells – Big Impact’ at 15:00 (UK) – 16:00 (CEST)
Join our eighth Free Friday chat group, featuring Dr Simon Saunders, Director – Technology, Real Wireless (http://www.realwireless.biz/).
Topic for discussion on Friday 10 May 2013 – ‘Small Cells – Big Impact’.
Subjects for this week:
- What are small cells?
– Classification of small cells (pico, metro etc)
– Why do mobile networks need small cells?
– Implications for small cells and LTE
– What are the technical innovations which have allowed small cells to emerge now?
– What is the state of the market?
– Prospects and challenges for the future?
– What does Cisco’s acquisition of Ubiquisys tell us about the state of the small cell market?
– What about inter-vendor interoperability?
– Small cells versus Wi-Fi?
– Backhaul?
About Dr Simon Saunders:
Simon is an independent specialist in the technology of wireless communications, with a technical and commercial background derived from senior appointments in both industry (including Philips and Motorola) and academia (University of Surrey). As co-founder of Real Wireless, he is responsible for overall technical capability and direction.
He is an author of over 150 articles, books and book chapters. He has acted as a consultant to companies including BAA, BBC, O2, Ofcom, BT, ntl, Mitsubishi and British Land and was CTO of Red-M and CEO of Cellular Design Services Ltd.
Simon speaks and chairs a wide range of international conferences and training courses and has invented over 15 patented wireless technologies. Particular expertise includes in-building wireless systems, radiowave propagation prediction, smart antenna design and mobile system analysis.
He has served on technical advisory boards of several companies, was Visiting Professor to the University of Surrey, member of the industrial advisory board at University College London and was founding chairman of Small Cell Forum (formerly Femto Forum), which he chaired from 2007-12.
He is currently a member of the Ofcom Spectrum Advisory Board and advisor to Quortus and Adjunct Professor at Trinity College Dublin.
It is our pleasure to have Simon with us on 10 May 2013.
Follow us on Twitter @ForumOxford – https://twitter.com/ForumOxford
Follow Real Wireless on Twitter @real_wireless – https://twitter.com/real_wireless
realwireless.biz realwireless.biz
Peter Holland • Ok – Good afternoon/evening or morning (as appropriate for your time zone) to everyone from the University of Oxford!
Welcome to ForumOxford’s FreeFriday chat on 10 May.
To those who are following the discussion, please feel free to post any questions during the discussion (refresh your screen regularly to see the latest comments)
Let’s begin. Ajit and Simon – over to you……
Ajit Jaokar • hi simon great to see you .. sorry linkedin was putting comments for review .. welcome!
Ajit Jaokar • ok lets start – first qs – Can you explain the basic ideas and motivation behind small cells
Simon Saunders • Thanks Ajit – great to be with you on one of my favourite wireless topics
Simon Saunders • For sure…small cells are something of a chameleon technology – all things to all people…
Simon Saunders • For some they are the best way to address a future ‘capacity crunch’ to meet the rising tide of mobile demand
Simon Saunders • …for others they are the best way to get great coverage to places macrocells cannot [economically] reach
Simon Saunders • For others still they are a tool for competition, to deliver enhanced quality of experience and a wider range of personalised services to their customers compared with their peers
Simon Saunders • There’s often confusion about what we do mean by small cells however. The short answer is – every base station in a mobile network which is not a macrocell.
Simon Saunders • Certainly including femtocells, picocells, and microcells in both their urban (metrocell) and rural (“meadowcell”?!) guises.
Ajit Jaokar • “every base station in a mobile network which is not a macrocell.” thats a good definition .. 
Ajit Jaokar • so whats the difference between femtocells, picocells, microcells and metrocells? are these being used interchangably?
Simon Saunders • The terms aren’t interchangeable, but likewise they don’t have hard boundaries. Femtocells = typically home and small office/SoHo. Picocells = Enterprise, retal etc (mainly indoors), Metrocells = busy cities, outdoors, “Meadowcells” = rural outdoor application of metrocells.
Ajit Jaokar • medowcells! first I heard that!
Simon Saunders • And for some they also include access points in unlicenced/licence exempt spectrum – i.e. mainly Wi-Fi. For other still they include every tech which puts the antennas close to the users, i.e. Distributed Antennas Systems and Cloud/Centralised-Ran solutions. But not everyone agrees with that!
Ajit Jaokar • Re “Distributed Antennas Systems and Cloud/Centralised-Ran solutions. But not everyone agrees with that!” – why not? ie where is the disagreement in the industry wrt it?
Simon Saunders • The disagreement on DAS and C-RAN is really about where each solution fits best, and on where the processing intelligence sits.
Simon Saunders • The original concept of small cells is to make use of cheap silicon processing at the edge, and keep the high-rate processing off the backhaul.
Simon Saunders • C-RAN allows the same radio benefits – antennas close to users – but centralises the processing. This allows each antenna unit to be closely coordinated with all the others, enabling exciting processing opportunities such as CoMP, but puts a big load on the backhaul.
Ajit Jaokar • @simon – thanks re c-ran concept
Simon Saunders • A word on DAS: a great workhorse, and currently the solution of choice for the biggest public venues and the largest enterprises.
Ajit Jaokar • thanks Simon. yes thats a good distinction. so next question – “What does Cisco’s acquisition of Ubiquisys tell us about the state of the small cell market?”
Simon Saunders • Cisco’s acquisition of Ubiquisys is a great validation of several things:
– the value of small cells
– the opportunity for small cells in the Enterprise
– the role of the UK in wireless innovation and tech innovation generally.
Ubi were definitely one of the key pioneers in the field – but many others deserve honourable mention too!
Ajit Jaokar • so whats the benefit ffor cisco? enterprise?
Simon Saunders • I look forward to the day when enterprise IT system integrators are as skilled at designing and implementing 3G/4G systems as they are today at implementing enterprise LAN/Wi-Fi: and are as accepted at that jon by mobile operators as their current SIs in the DAS space.
Ajit Jaokar • ok next qs – do small cells and wifi complement each other? (or whats your view)
Simon Saunders • Great question – and the industry has changed on this point (small cells and Wi-Fi) over time.
2005-7: Some think femtocells could be the mobile competitive response to Wi-Fi…
Simon Saunders • 2007-2009: Femtocells and Wi-Fi complement each other: different characteristics, different devices, different places…
2010 – today: If we implement both small cells and Wi-Fi in a coordinated fashion it’s better for operators and consumers alike. And we need to use all the spectrum we can get!
Ajit Jaokar • @simon – thanks(re wifi) yes we need all spectrum
Dr John Kelliher IKC • Surely the concept of edge is somewhat of a misnomer – an artificial concept that should not be introduced or exist in cellular ? Well, why should it ?
Ajit Jaokar • we will take @John’s qs now re concept of edge?
Simon Saunders • @John:
The question of where is “the edge” is a good one. I was using it earlier to mean simply “close to the user”. But there are many options for how to distribute processing across a network and these options suit different use cases. C-RAN and femtocells represent two extremes of those options, but there are many useful cases in between which deserve more attention.
Ajit Jaokar • @thanks simon (re edge insights)
Ajit Jaokar • coming back to Operators .. other than spectrum case.. Why do mobile networks need small cells?
Simon Saunders • Mobile operators need small cells because:
– You can’t multiply today’s macrocells by x10
– Spectrum efficiency is limited
– Space (and money) for antennas on macros and in devices is limited
– User demand higher rates, reliably – and even without considering capacity, macrocell geometries struggle with that
…
Simon Saunders • …and operators need small cells to remain different. One day operators will advertise that they provide the best service because they have the most small cells. That may seem implausible today, but watch this space!
Ajit Jaokar • @simon – very insightful ” One day operators will advertise that they provide the best service because they have the most small cells. That may seem implausible today, but watch this space!”
Dr John Kelliher IKC • @Simon – Cisco and Juniper have a somewhat traditional view of Internetwork edge, which I thought you may be referring to. However processing should, I would be consider be a distributed function, granted that the air-interface processing is high. Operators usually consider use cases as multifaceted { end user – designer/implementer – Architectural }, so ordinarily I would consider this to be more a design consideration than a use case.
Simon Saunders • @John: I think the best location of the processing is indeed a design consideration, but the optimum location is a function of the use case.
Simon Saunders • .. so the best processing location for a remote rural village is entirely different to that for a national football stadium!
Ajit Jaokar • @simon – What about inter-vendor interoperability?
Simon Saunders • Interoperability = big topic.
Starting point: femtocells needed standards to get off the ground. And got them very early in the lifecycle I’m proud to say,
Simon Saunders • But interop is much more than standards. We organised plugfests to encourage and demo actual interop and to iron out standards ‘bugs’. But even then you have to advance step by step towards meaningful functionality. Today operators can genuinely buy femtocells from different vendors and have them work with a third vendor’s gateway, though there is still more pain than I’d like in that process. Nevetheless for femtocells it’s all trending in the ‘right’ direction.
Ajit Jaokar • and the femto forum had a role to play
Ajit Jaokar • @simon yes agree tending in right direction ..
Simon Saunders • Metrocells are in a different state. They need very tight integration with the surrounding macro network. And there are sharp compromises at this stage between interop and performance. And tough choices for operators who want one “neck to choke”!
Simon Saunders • And let me be clear: open interop is not only good for operators, but also for vendors who want to address a wide market, and also for consumers who want products which work well. But it has to be done without sacrificing innovation or performance unduly,
Ajit Jaokar • @simon – thanks for insights on macro cells. agree re innovation
Ajit Jaokar • from a custromer standpoint femtocells have a cost?
Simon Saunders • For consumers, femtocells bring benefits which I believe outweigh costs. If you are genuiniely needy, most operators today will let you have a femtocell as part of the package (in exchange for a little loyalty!).
Simon Saunders • Another way of looking at it: given femtocells can (in the right circumstances) reduce overall costs of delivering a service for the operator, and eventually the customers pay for every cost in the network, reduced cost for the operator translates to reduced cost for the customer (assumes a competitive market!)
Ajit Jaokar • yes agree. win-win for all
Ajit Jaokar • ok we are in last 15 mins .. @simon anything imp we have not covered you think?
Simon Saunders • A tangential topic: for those considering research in this area, we need:
– More analysis of the lifetime environmental & energy impact of small cells
– Radio techniques for hyperdense networks
Simon Saunders • Would also like to give due acknowledgement to the >145 organisations (operators and vendors) in Small Cell (Fermto) Forum who made all this happen. When we started it was heresy – now 98% of operators want small cells in their networks. Prevailing orthodoxy!
Simon Saunders • But (maybe finally) there are unanswered questions: exactly how many small cells are needed, and exactly where they should go, and exactly how they are managed – are all open questions. We need to rewrite the mobile network planning rulebook!
Ajit Jaokar • @simon – yes indeed – great work by femto forum and great to have you here for that reason
Simon Saunders • Thanks @Ajit and all online – great to chat. Go and ask your mobile operator for a femtocell!
Ajit Jaokar • indeed! thats a good point to conclude!
Peter Holland • Thank you to Ajit and Simon for a great discussion and also thank you to all those of you who followed and contributed to the conversation.
It’s goodbye from ForumOxford at the University of Oxford for now.
Event: Aligning Small Cell Incentives
The 