Next generation thinking Examining small cell backhaul requirements 15 February 2012
Examining small cell backhaul requirements Agenda 5 mins 15 mins 10 mins 10 mins Are small cells really the next big thing? Lance Hiley, VP Marketing, Cambridge Broadband Networks Ltd The challenges How will operators deploy small cells? Key design considerations for small cell backhaul Julius Robson, Wireless Technology Consultant and Leader, NGMN Small Cell Backhaul Requirements Group The solutions How do different solutions compare against the requirements? Lance Hiley Your questions Q&A open for 10 minutes 2
Who we are Formed in 2000 Global marketshare leader in line of sight multipoint microwave technology Suitable for LTE network backhaul Selling to 7 of the top 10 mobile operator groups 3
Are small cells really the next big thing? Lance Hiley VP Marketing Cambridge Broadband Networks Limited 4
Are small cells really the next big thing? 1% smartphone users consume 50% of mobile data (what happens when others catch on?) More recent and realistic version of Cisco VNI still shows growth New devices and apps will use whatever capacity is available Industry is organising itself to speed small cells to market 5
Small cells could be the answer Mobile cellular networks were initially designed for voice Best Signal Quality in Cellular Networks: Asymptotic Properties and Applications to Mobility Management in Small Cell Networks, Alcatel-Lucent, 2010 http://jwcn.eurasipjournals.com/content/2010/1/690161 The popularity of mobile broadband multimedia services has redefined the RAN and backhaul requirements of mobile networks: data is dominant Mobile networks have to evolve to transport packet data traffic efficiently: data is different Reducing cell size is one of the most effective ways to improve the spatial reuse of radio resources and increases network capacity Bringing bandwidth closer users improves customer quality of experience 6
Small cells could be the answer Small cells can ease congestion in busy areas by serving hot spots and indoor users, leaving macro-layer to deal with wide-area high-mobility outdoor users In this webinar we consider the implications of this trend on the backhaul 7
The challenges: How will operators deploy small cells? Resulting requirements for small cell backhaul Julius Robson Wireless Technology Consultant Leader, NGMN Small Cell Backhaul Requirements Group 8
Why deploy small cells? for Hot spots and Not spots macro Easing congestion within macro coverage New coverage in addition to macro A small cell will improve both coverage and capacity, but the primary motive is important when considering backhaul requirements 9
Where will they be? Congestion on fully upgraded macro sites Need to densify No rooftop space left smaller units needed to fit available locations Small cells Smaller unit = less power = shorter range Small, low power cells close to users Near street level Small cell sites typically 4-6 m above street level, on sides of buildings or street furniture 10
Case study: what density of small cells is needed? Case study of how demand density will be supplied with a mix of HSPA, LTE and small cells Gives site densities and spacing 5 sites/km 2 dense macro rooftop network Small cells exceed this in ~2013, requiring below rooftop Spacing will be lower than average in pockets of high demand ~100-200m Assumptions Demand growth from PA consulting 1 Spectral efficiency evolution Ofcom 2 Macro site density 5/km2 (Holma 3 ) 1 Predicting areas of spectrum shortage, PA Consulting, April 2009 2 "4G Capacity Gains", Real Wireless for Ofcom, Dec 2010 3 LTE for UMTS: Evolution to LTE Advanced, Harri Holma, Wiley 2010 Variation due to non uniform deployment Dense macro 11
The what and how of backhaul requirements 1) Fundamentals What Coverage Capacity Cost Architecture Small Cell Backhaul Solution 2) Practicalities How Size & weight Spectrum bands Integration Installation Backhaul features (QoS, Sync etc) Availability/latency Implementation 12
The backhaul coverage challenge PoP Small Cells 13
Backhaul coverage requirements Coverage from: Points of Presence PoP locations: e.g. rooftop macrosites PoPs density ~5 sites /km 2 PoP Coverage to: Small cell sites Locations:4-6m above street level Densities: increasing over time Estimate 30 sites per km 2 ~100-200m spacing in areas of high demand PoP Coverage = Connectivity between PoP and small cell sites with sufficient QoS 14
Quality of Service over Backhaul Operators want consumer QoE to be independent of the access topology Backhaul QoS should be driven by services offered Some aspects of backhaul QoS may change according to deployment scenario: Aspect of backhaul QoS Small cell deployed primarily for New coverage @Not Spot Easing congestion @Hot Spot Availability same as macro relaxed Delay (Latency, jitter) same as macro same as macro Capacity provisioning relaxed greater than small cell Where easing congestion, RAN capacity should not be limited by the backhaul Where coverage overlaps, macro layer acts as fall back for small cells Macrocells might be quality not quantity.but the reverse is not true for small cells 15
? Backhaul capacity provisioning HSDPA 2x2 64 QAM DC HSDPA 2x2 64 QAM LTE 10MHz 2x2 LTE 20MHz 2x2 6 42 Loaded Peak 12 18 84 75 34 150 0 50 100 150 200 DL Capacity Provisioning per small cell, Mbps Assumptions Modified version of NGMN s macrocell backhaul capacity provisioning [1,2] Includes user plane traffic plus overheads for transport, X2 and IPsec Loaded macrocell throughputs scaled by 125% according to 3GPP simulations [1] "Guidelines for LTE Backhaul Traffic Estimation", NGMN Alliance, July 2011, http://goo.gl/ewqqg [2] NGMN Alliance Optimised backhaul solutions for LTE, challenges of Small Cell deployment and Coordinated QoS, NGMN Alliance, Layer 123 LTE/EPC & Converged Mobile Backhaul, December 2011 [3] "Further advancements for E-UTRA physical layer aspects", 3GPP TR 36.814 V9.0.0 (2010-03) Loaded figure represents busy times. Peak represents maximum capability of the RAN during quiet times Small cell sites will initially be single carrier, single cell and single generation, hence need less backhaul capacity than multi-sector, carrier and operator macros This reduces on site aggregation gains so backhaul traffic will be burstier 16
Backhaul cost requirements $ TCO per site Capex Opex Equipment Installation Site rental Power Last mile backhaul Maintenance RAN backhaul leased line spectrum etc Cost per bit is likely to be similar to that of macro sites, but many small cells will be needed to supply same capacity as a macro so cost per small cell site will need to be much lower 17
Physical design requirements The small cell and backhaul unit combined should be Small enough to fit in available street level locations Planning/zoning may impose volume/dimension restrictions Lightweight to facilitate installation A one man lift & mount can reduce costs Environmental Innocuous rather than sexy Should not draw attention to itself Touch safe and tamper proof Some sites may be within reach of the public Size Power Reliability? Appearance Planning Permission Connectivity Weight Installation & Commissioning Backhaul/RAN integration 18
How do different solutions compare? Lance Hiley VP Marketing Cambridge Broadband Networks Limited 19
Small cell backhaul options Conventional PtP For: High capacity Against: Coverage awkward, spectrum opex, high installation costs E-band For: High capacity Against: High capex and opex Fibre (leased or built) For: High capacity (if you pay enough) Against: Recurring charges, availability and time to deploy Non-line of sight multipoint microwave For: Good coverage, low cost of ownership Against: Low capacity, spectrum can be expensive 20
How does it all connect up - wirelessly Tree (point-to-point) Ring Mesh Multipoint Key small cell pop Links low capacity high capacity with redundancy 21
Point-to-Point (PtP) microwave PtP Microwave Lots of bandwidth microwave frequencies available at 10-60GHz but oversubscribed in many urban centres PtP spectrum is link-licensed; high recurring opex Area licensing can address this when available PtP links use two radios: each requiring space, installation, energy: high recurring opex PtP E-band 10GHz of spectrum available at 71-76 and 81 GHz a window between peaks of high atmospheric absorption Light licensing conditions reduces spectrum opex in many markets Installation of equipment is trickier than conventional PtP PtP The most common microwave topology For N links, 2N radios Dedicated RF channel for each node B served Well-suited to constant bit rate traffic Well-suited to long links Conventional and E-Band frequencies Multiple radios, antenna s per site to support ring/mesh topologies makes PtP difficult to deploy at street level 22
Fibre Fibre Great where already available, otherwise slow and costly to install High-capacity, low-latency connection High recurring cost even in competitive markets UK published fibre pricing 34 mbps 140 mbps 280 mbps 500 mbps Installation $ 2,000 $2,000 $2,000 $2,000 Yearly rental fees $10,000 $14,000 $20,000 $30,000 23
Non-line of sight (NLoS) microwave Good for coverage, capacity limited by available spectrum NLoS propagation requires low carrier frequencies prized for mobile access itself Free spectrum worth every penny...but Wi-Fi uses the entire unlicensed low frequency spectrum Spectral efficiency advances unlikely to compensate: access and backhaul operating in same (NLoS) environment Unpaired TDD spectrum could be used for NLoS backhaul, but quantity of is small compared to the LTE and HSPA bands it has to backhaul The 3.5 GHz band is large and underused, however 3GPP is planning UMTS (HSPA) and LTE specifications 24
Line of sight (LoS) multipoint microwave Multipoint microwave designed for streetlevel deployment High-capacity multipoint microwave operating at ETSI PMP frequencies: 10.5, 26 and 28GHz. Other bands in consideration Backhaul 8 remote terminals per access point with up to 300Mbps backhaul capacity Integrated antenna for maximum deployment flexibility/lowest operational cost Multipoint microwave: fastest growing microwave topology today For N links, N+1 radios Shared RF channel amongst all sites Well-suited to variable bit rate (bursty) traffic Well-suited to dense environments Spectrum under-subscribed in most markets Point-to-Multipoint (PMP) aggregates packet traffic from multiple RT s Uses 40% less spectrum Only one radio per small cell site 25
Small cell backhaul revolution PMP hubs beam high-capacity multipoint bandwidth down urban canyons Large numbers of links for small cells, with high peak to average data traffic favour PMP aggregation capabilities 26
PMP best fit across small cell backhaul requirements LoS PTP and eband requirement of two radios per link impacts equipment/installation costs NLoS wireless capacity is limited Leased line connections have high repetitive costs Wi-Fi range compromises backhaul application 27
Cost per Mb/s traffic carried Architecture contributes to lowering cost of transport 9,000 8,000 7,000 6,000 5,000 4,000 3,000 Small Cell TCO (Capex & Opex) As traffic builds on a small cell network, cost of transport drops with all solutions (blip seen for fibre caused by transitioning to higher-capacity service) Multipoint architecture delivers lower cost of transport sooner - from the moment of installation 2,000 1,000 0 32 Mb/s 80 Mb/s 120 Mb/s 150 Mb/s Fiber, leased Eband PTP PMP Expon. (PMP) 28
Summary Operators need high-capacity, low-opex backhaul for small cell network densification Small cells needed to supply Hot Spots and densify network, offloading macro for high-mobility users Multipoint LoS microwave is a mature technology option for backhaul: High-capacity Short deployment time Low cost of ownership Spectrum readily available Cambridge Broadband Networks VectaStar Metro meets the small cell backhaul challenge Read our whitepaper: http://cbnl.com/resources/white-papers 29
Your questions Lance Hiley: lhiley@cbnl.com Julius Robson: jrobson@cbnl.com Download the white paper: http://cbnl.com/resources/white-papers Copyright Cambridge Broadband Networks Limited. All rights reserved.