Business feasibility analysis of use of TV WS for mobile broadband services taking into account spectrum pricing

Transcription

1 Business feasibility analysis of use of TV WS for mobile broadband services taking into account spectrum pricing Jan Markendahl and Pamela Gonzalez Sanchez Royal Institute of Technology (KTH) Bengt G. Mölleryd Swedish Post and Telecom Agency (PTS) Stockholm, Sweden COST-TERRA meeting, Brussels, November 15-17, 2011

2 Topics today Amount of spectrum per operator The Multitude of business and services scenarios Capacity, cost and cost structure Principles and trends, Example results of cost savings Impact of spectrum costs Countries are different 2

3 Research questions RQ1. What kinds of business opportunities can be identified for different situations where diffrent types of mobile operators make use of secondary spectrum access? RQ2. Is there a common business feasibility analysis method or do we need a multitude of case dependent methods Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

4 Why do operators want spectrum? 4

5 Why do operators want spectrum? More bandwidths means: Higher data rates: better offer competitive advantage Data rates ~ fixed broadband? Higher capacity: more traffic, more users can be served More cost-efficient deployment 5

6 Why do operators want spectrum? High bandwidth means high capacity per site, i.e less number of base station sites Licensed Licensed Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

7 Why do operators want spectrum? High bandwidth means high capacity per site, i.e less number of base station sites Licensed White space Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

8 Number of sites Impact of amount of spectrum Case: New urban office area with many users Numner of base station sites Used spectrum (MHz) Case: Traffic from office workers In 1 km2 area Spectral eff = 1,70 (LTE type) Spectral eff = 0,70 (HSPA type) 8

9 Number of sites Impact of amount of spectrum Case: New urban office area with many users Numner of base station sites Used spectrum (MHz) Case: Traffic from office workers In 1 km2 area Spectral eff = 1,70 (LTE type) Spectral eff = 0,70 (HSPA type) Cases for different Swedish operators using 2.6 GHz band 9

10 Expected MBB traffic increase: >1000 times More spectrum: < 2 times By Mölleryd & Markendahl from ITS Conf 2011 Valuation of spectrum for mobile broadband services - Engineering value versus willingness to pay Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

11 Number of available TV channels in Sweden from Quasar deliverable 5.1 TV WS 1 TV channel ~ 8 MHz 5 TV channels ~ 40 MHz Compare with LTE deployment in 800 MHz and 2.6 GHz bands 2.6 GHz Telia, Tele 2 and Telenor have 2* 20 MHz - 3 has 2*10 MHz x Telia and 3 have 2*10 MHz - Telenor and Tele2 joint venture (Net4Mobility) has 2*10 MHz Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16, 2011 x

12 A Multitude of Multitudes - Scenarios & Sub-scenarios The business viability is not an inherent property of any type secondary use or the type of primary spectrum The business viability depends on specific business cases incl. services, demand and types of network deployment AND the business cases of competiting solutions Each service scenario consists of multiple sub-scenarios The availability of spectrum for secondary use The type of service intended for secondary use The demand and willingness to pay for the service The cost structure for network deployment options Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

13 A Multitude of Multitudes - QUASAR Service Scenarios Cellular Use of White Spaces WiFi-like Use of White Spaces Secondary Wireless Backhaul Secondary Spectrum Commons in Radar Band Indoor Broadband in Aeronautical Spectrum Cognitive Machine-to-Machine (Infrastructured) Cognitive Machine-to-Machine (Ad hoc) Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

14 A Multitude of Multitudes - QUASAR Service Scenarios MBBA in rural areas using TV WS only Cellular Use of White Spaces WiFi-like Use of White Spaces MBBA in urban areas using TV WS only MBBA in urban areas using TV WS as complement MBBA using TV WS in indoor locations Secondary Wireless Backhaul Secondary Spectrum Commons in Radar Band Indoor Broadband in Aeronautical Spectrum Cognitive Machine-to-Machine (Infrastructured) Cognitive Machine-to-Machine (Ad hoc) Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

15 Summary of findings you can see different things Provisioning of MBBA in rural areas using TV WS - TV white space spectrum is available - The demand is low, primary spectrum is not fully utilized MNO using TV WS for capacity expansion in urban areas - Few channels are available, still -20 MHz can be found - TV WS used as complement reduces the need to build new sites An operator without own spectrum using TV WS and own infrastructure in order to offer MBB services in urban areas - cost structure of new sites results in (too?) high costs A local operator offering indoor capacity to mobile operators - TV WS availability is higher for indoor low power deployment - Business cases supported by trend to share indoor infrastructure Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

16 Summary of findings you can see different things Provisioning of MBBA in rural areas using TV WS - TV white space spectrum is available - The demand is low, primary spectrum is not fully utilized MNO using TV WS for capacity expansion in urban areas - Few channels are available, still -20 MHz can be found - TV WS used as complement reduces the need to build new sites An operator without own spectrum using TV WS and own infrastructure in order to offer MBB services in urban areas - cost structure of new sites results in (too?) high costs A local operator offering indoor capacity to mobile operators - TV WS availability is higher for indoor low power deployment - Business cases supported by trend to share indoor infrastructure Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

17 Topics today Amount of spectrum per operator The Multitude of business and services scenarios Capacity, cost and cost structure Principles and trends, Example results of cost savings Impact of spectrum costs Countries are different 17

18 Capacity, cost and cost structure of HSPA and LTE Total cost for new site ~ 200k Radio Total cost for new site ~ 110k Assuming 3 sector site 20 MHz of bandwidth and cell average spectral efficiency 0,7 bps per Hz (HSPA) 1,7 bps per Hz (LTE) Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans HSPA year 2008 LTE year 2010 LTE year

19 Capacity, cost and cost structure of operator business cases Total cost for new site ~ 200k Radio Total cost for new site ~ 110k Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Existing MNO 19

20 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Existing MNO 20

21 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Existing MNO 21

22 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Existing MNO Nov 16,

23 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Existing MNO Nov 16,

24 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Assuming 3 sector site 20 MHz of bandwidth and cell average spectral efficiency 0,7 bps per Hz (HSPA) 1,7 bps per Hz (LTE) Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans TV WS + new site LTE +new site TV WS+ re-used site Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

25 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans TV WS + new site LTE +new site TV WS+ re-used site Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

26 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k?? Radio Trans TV WS + new site LTE +new site TV WS+ re-used site Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

27 Work in progress example of type of presentation Deployment Costs [KEuros] URBAN Scenario "EUROPE CASE" Urban LTE Greenfield Urban TVWS Greenfield Urban LTE Incumbent Urban TVWS Incumbent Costs reduction Demand [Mbps/Km 2 ] 27

28 YGODOWYPI 28

29 YGODOWYPI You Get Output Depending On What Yo Put In 29

30 YGODOWYPI You Get Output Depending On What Yo Put In Radio Radio Trans missio n Sites Trans missio n Sites 30

31 YGODOWYPI You Get Output Depending On What Yo Put In Cost of spectrum cost? Radio Radio Trans missio n Sites Trans missio n Sites 31

32 Capacity, cost and cost structure of operator business cases Total cost for new site ~ 200k Radio Total cost for new site ~ 110k YGODOWYPI Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Existing MNO Nov 16,

33 Spectrum price in relation to other cost?? Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

34 Prices paid for spectrum (by Mölleryd ITS 2011) Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

35 Prices paid for spectrum (by Mölleryd ITS 2011) India/Dehli Bahrti 5 MHz Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

36 Spectrum price in relation to other costs - normalized to auction cost per sites 10 k radio ~100k Site + trans radio trans 10 k ~20 k Deploy new site Upgrade existing sites Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

37 Spectrum price in relation to other costs - normalized to auction cost per site > 100 k 10 k radio ~100k Site + trans ~15 k ~0,10 k radio trans 10 k ~20 k Netherlands Tele2 2.6 GHz Sweden Telia 800 MHz India Bahrti/ Dehli Deploy new site Upgrade existing sites Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

38 Network costs depending on bandwidth Deployment Costs [KEuros] URBAN Scenario "EUROPE CASE" Urban LTE Greenfield Urban TVWS Greenfield Urban LTE Incumbent Urban TVWS Incumbent Bandwidth [MHz] 38

39 Network costs depending on bandwidth 1500 URBAN Scenario "INDIA CASE" Deployment Costs [KEuros] We observe the effect of high spectrum costs Urban LTE Greenfield Urban TVWS Greenfield Urban LTE Incumbent Urban TVWS Incumbent Bandwidth [MHz] 39

40 Topics today Amount of spectrum per operator The Multitude of business and services scenarios Capacity, cost and cost structure Principles and trends, Example results of cost savings Impact of spectrum costs Countries are different 40

41 Prices paid for spectrum (by Mölleryd ITS 2011) India/Dehli Bahrti 5 MHz Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

42 Countries in the world are different Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

43 Countries in the world are different 43

44 44

45 Conclusions Provisioning of MBBA in rural areas using TV WS MNO using TV WS for capacity expansion in urban areas An operator without own spectrum using TV WS and own infrastructure in order to offer MBB services in urban areas A local operator offering indoor capacity to mobile operators Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

46 Conclusions Provisioning of MBBA in rural areas using TV WS MNO using TV WS for capacity expansion in urban areas An operator without own spectrum using TV WS and own infrastructure in order to offer MBB services in urban areas A local operator offering indoor capacity to mobile operators Dig more Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

47 Conclusions Provisioning of MBBA in rural areas using TV WS MNO using TV WS for capacity expansion in urban areas An operator without own spectrum using TV WS and own infrastructure in order to offer MBB services in urban areas A local operator offering indoor capacity to mobile operators Look for an analyze TV WS business opportunities in a non-european context - Less amount of licensed spectrum - High spectrum prices - More chaotic and un-coordinated markets Dig more Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

48 Thanks for your attention My Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

49 Base station site locations in urban areas from PTS Transmitter map web page, Downtown Stockholm Kista Industry Area 49

50 Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels

51 Offered bit rate vs coverage & load promised data rate at low level promised data rate at higher level 51

52 Offered bit rate vs coverage & load promised data rate at low level promised data rate at higher level 52

53 Data rate, bandwidth and aggregation of carriers or bands The higher bandwidth the higher the data rate 53

54 Data rate, bandwidth and aggregation of carriers or bands The higher bandwidth the higher the data rate TV WS combined with 800MHz 54

55 Example peak data rate Operators with different amount of spectrum - One operator using 10 MHz in one band - One operator using 10 MHz in two bands, in total 20 MHz - One operator using 20 MHz in two bands, in total 40 MHz - Two cooperating operators each using 20 MHz in two bands, in total 80 MHz 55

56 From here slides in old file 56

57 Prices paid for spectrum (by Mölleryd ITS 2011) Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

58 Conclusions BUT! Also consider other solutions - WLAN - Licensed > 3 GHz You need to compare solutions using TV WS with existing solutions and business cases ; i.e. the offers and cost structure of competitors Example: Using TV WS for Mobile broadband New actors with TV WS spectrum only offering wide area mobile broad band access services will have problems Local operators can use TV (or radar) WS to offer indoor access Exploit benefits of TV WS for existing solutions and actors Capacity => delay or reduce need for deployment of new sites Data rate=> with aggregation higher data rates can be offered Cost? => Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

59 The dead zone problem Femtocell base station cause interference in the downlink for UEs connected to macro base stations, i.e. for closed access Areas without any coverage (dead zones, coverage holes) appear in the vicinity of the femto base stations 59

60 The dead zone problem Occurs for closed acess femtocells Dead zones are largest at cell borders ~ 20 m for co-channel operation (WCDMA) 1-2 m for adjacent channels (WCDMA) 60

61 Macrocell-femtocell interference Macro BS Femto BS Macro BS Small separation Femto BS Co-channel operation Adjacent channel with guard band Macro BS Femto BS Macro BS Large separation Femto BS Adjacent channel operation Dedicated femtocell band 61

62 Cost and capacity for macrocellular and femtocell deployment Total cost ~ 110k Radio Site capacity ~100 Mbps Site And Trans Site capacity ~100 Mbps Cost for upgrading existing site ~ 20k Radio Trans Capacity per access point Mbps Deployment of Upgrading of Deployment of new macro sites existing macro sites Femtocells Cost per access point ~ 1 k 62

63 Local operator using TV WS to offer indoor capacity A local operator can deploy a femtocell network where only a fraction of the femtocell capacity is used A few available MHz can satisfy even very high demand levels Example 1: One 5 MHz femtocell with Spectral eff = 8 bps/hz can capacity wize serve GB per month users Example 2: 1 MHz is enough to serve 40 users (with spectral effiency =4 bps/hz) => One 8 MHz channel can be used for several 1 MHz femtocells This makes all bands and also a few MHz of white space interesting 63

64 Conclusions You need to compare solutions using TV WS with existing solutions and business cases ; i.e. the offers and cost structure of competitors Example mobile using TV WS for Mobile broadband New actors with TV WS spectrum only offering wide area mobile broad band access services will have problems Local operators can use TV (or radar) WS to offer indoor access Exploit benefits of TV WS for existing solutions and actors Capacity => delay or reduce need for deployment of new sites Data rate=> with aggregation higher data rates can be offered 64

65 Conclusions BUT! Also consider other solutions - WLAN - Licensed > 3 GHz You need to compare solutions using TV WS with existing solutions and business cases ; i.e. the offers and cost structure of competitors Example mobile using TV WS for Mobile broadband New actors with TV WS spectrum only offering wide area mobile broad band access services will have problems Local operators can use TV (or radar) WS to offer indoor access Exploit benefits of TV WS for existing solutions and actors Capacity => delay or reduce need for deployment of new sites Data rate=> with aggregation higher data rates can be offered 65

66 Conclusions You need to compare solutions using TV WS with existing solutions and business cases ; i.e. the offers and cost structure of competitors Example mobile using TV WS for Mobile broadband New actors with TV WS spectrum only offering wide area mobile broad band access services will have problems Local operators can use TV (or radar) WS to offer indoor access Exploit benefits of TV WS for existing solutions and actors Capacity => delay or reduce need for deployment of new sites Data rate=> with aggregation higher data rates can be offered Cost? => 66

67 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Assuming 3 sector site 20 MHz of bandwidth and cell average spectral efficiency 0,7 bps per Hz (HSPA) 1,7 bps per Hz (LTE) Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Existing MNO 67

68 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Assuming 3 sector site 20 MHz of bandwidth and cell average spectral efficiency 0,7 bps per Hz (HSPA) 1,7 bps per Hz (LTE) Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Existing MNO 68

69 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Assuming 3 sector site 20 MHz of bandwidth and cell average spectral efficiency 0,7 bps per Hz (HSPA) 1,7 bps per Hz (LTE) Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k?? Radio Trans New TV WS operator Greenfield MNO Existing MNO 69

70 70

71 Thanks for your attention My Link to my PhD Thesis, February 2011 Mobile Network operators and cooperation A tele-economic study of infrastructure sharing and mobile payment services 71

72 MNO using TV WS as complement Consider office areas and compare macrocell deployment and Femtocell (indoor) deployment An operator with existing sites would prefer macrocell deployment rather than femtocell deployment due to cost reasons Many femtocells are needed to ensure coverage Femtocell deployment (currently) leads to overprovisining of capacity This is an opportunity for local operators (LTE) TV WS femtocell deployment can be one way to avoid co-existence problems that occur when macro and femtocells use the same band, e.g. dead zones 72

73 Ongoing work - Use of TV WS for mobile broadband access Summary of initial results you can see different things Provisioning of MBBA in rural areas - TV white space spectrum is available - The demand is low, primary spectrum is not fully utilized MNO using TV WS for capacity expansion in urban areas - Few channels are available, still -20 MHz can be found - TV WS used as complement reduces the need to build new sites An operator without own spectrum using TV WS and own infrastructure in order to offer MBB services in urban areas - cost structure of new sites results in (too?) high costs A local operator offering indoor capacity to mobile operators - TV WS availability is higher for indoor low power deployment - Business cases supported by trend to share indoor infrastructure 73

74 MNO using TV WS as complement Consider office areas and Macrocell deployment Femtocell (indoor) deployment An operator with existing sites would prefer macrocell deployment rather than femtocell deployment due to cost reasons Many femtocells are needed to ensure coverage Femtocell deployment (currently) leads to overprovisining of capacity This is an opportunity for local operators (LTE) 74

75 General about the methodology The analysis is done for specific sub scenarios Links are established between the technical performance and the business case Key characteristics of the cost structure are identified and included in the analysis We can do the analysis without knowledge about the exact availability of spectrum, bandwidth, distribution in time and space We start to do a what IF analysis (sensitivity analysis) When detailed technical data (spectrum availability and performance) are available we can include this 75

76 Bandwitdh and spectral efficiency The available frequency channel has a bandwidth In the general case some average width and some variation in time and space Frequency The frequency channel has a spectral efficiency (SE) In the general case some average SE and some variation in time and space 76

77 Spectral efficiency We will assume an equivalent usable bandwidth and an average spectral efficiency This will be used as the resource used for dimensioning of the wireless capacity Bandwidth SE 77

78 Compare Spectral efficiency for cellullar systems Peak data rate ~10-20 bps per Hz Cell border rate < 0,10 bps per Hz Average data rate ~1-2 bps per Hz Average cell border Distance from base station 78

79 The spectral efficiency depends on The type of system used for secondary usage The deployment and location of primary and secondary system Bandwidth SE Availability data - squares 79

80 The spectral efficiency depends on The type of system used for secondary usage The deployment and location of primary and secondary system x No available TV channels per square x

81 Compare Spectral efficiency for cellullar systems Peak data rate ~10-20 bps per Hz Cell border rate < 0,10 bps per Hz Average data rate ~1-2 bps per Hz As a first step approximation this average value representative for the whole usage area is used Sensitivity analysis can be made with higher/lower values of spectral efficiency Average cell border 81

82 First approximation model With this description we can model Sharing of bands between multiple actors Variations in availability, interference levels or signal quality SE SE Bandwidth 82

83 First approximation model With this description we can model Sharing of bands between multiple actors Variations in availability, interference levels or signal quality SE SE Bandwidth Bandwidth 83

84 From British Telecom paper presented at ISART 2010: TVWS band provides coverage similar to a mobile broadband network with a 20% deployment density 84

85 Number of available TV channels in Sweden Figure 4.14 in Quasar deliverable TV channel ~ 8 MHz 5 TV channels ~ 40 MHz Compare with LTE deployment in 800 MHz band with bandwidth ~10 MHZ x x

86 Capacity, cost and cost structure of operator business cases Total cost for new site ~ X 200k Radio Total cost for new site ~ 110k Assuming 3 sector site 20 MHz of bandwidth and cell average spectral efficiency 0,7 bps per Hz (HSPA) 1,7 bps per Hz (LTE) Capacity of radio base station site ~ 40 Mbps Site and Trans Capacity of radio base station site ~ 100 Mbps Radio Site And Trans Capacity of radio base station site ~ 100 Mbps Cost for upgrading an existing site ~ 30k Radio Trans New TV WS operator Greenfield MNO Existing MNO 86

87 Radio capacity Radio capacity Data Rate Data Rate 87

88 Peak data rate is very important Marketing by swedish mobile operators Compare 3G and 4G 88

89 Femtocells 89

90 Different solutions for indoor coverage DAS: Distributed Antenna System Indoor base stations Repeater Σ Base station Red MNO Base station Blue MNO 90

91 Wireless indoor solutions Indoor solutions are not only used in order to compensate for wall penetration losses Other reasons may be: Companies want ensured and dedicated capacity Companies use mobile phones as office phone Mobile operators want to increase customer loyalty Mobile operators want to offload data traffic from outdoor (more expensive?) macro networks 91

92 Indoor wireless solutions are used in two different business settings B2B2C 1. To ensure public access in shopping malls, subways, sport arenas The users are subscribers of the operators that visit the shopping mall, etc The operators have agreements with the owners of the mall, the arena, etc The service IS the ensured indoor coverage B2B 2. To provide indoor private access at company offices etc as part of a complete offer The users are the employees of the company, etc The indoor coverage is just one part of the offer Other components can be outdoor coverage, handsets, IT support and services, call centers 92

93 Options for femtocell deployment Parallell single-operator networks Less feasible due to cost and operation Multi-operator access points Cheap? not on 3GPP agenda Single operator femtocell network One multi-operator network Based on national indoor roaming 93

94 Shared indoor infrastructure Red Operator Blue Operator Payments and billing Customer rel mgmt Payments and billing Customer rel mgmt Facility owner Shared indoor infrastructure 94

95 Business feasibility of femtocells Femtocells are feasible for deployment at offices Scalable solution, access for a specific user group Open issues for femtocells deployed for public access Difficult with multiple femtocell networks Cost benefits compared to macro networks is unclear Interference problems and spectrum allocation issues 95

96 Example of deployment costs for indoor demand 4 Deployment costs Macronew sites Network CAPEX ( M )s 3,5 3 2,5 2 1,5 1 0,5 Femto - 4 users Macroreused denser_5_times DAS Macro-reused Capacity (Gbps) Femto-32 users 96

97 Business opportunity Business opportunity depends on the specific scenario / use case The type of service The user density, demand and willingness to pay The type of radio network deployment If the spectrum is used as the only resource or as a complement The business cases of competiting solutions 97

98 A Multitude of Multitudes - QUASAR Service Scenarios Cellular Use of White Spaces WiFi-like Use of White Spaces Secondary Wireless Backhaul Secondary Spectrum Commons in Radar Band Indoor Broadband in Aeronautical Spectrum Cognitive Machine-to-Machine (Infrastructured) Cognitive Machine-to-Machine (Ad hoc) 98

99 A Multitude of Multitudes - QUASAR Service Scenarios Cellular Use of White Spaces MBBA in rural areas macrocell deployment MBBA in urban areas macrocell deployment WiFi-like Use of White Spaces MBBA in urban areas - femtocell deployment Secondary Wireless Backhaul Secondary Spectrum Commons in Radar Band Indoor Broadband in Aeronautical Spectrum Cognitive Machine-to-Machine (Infrastructured) Cognitive Machine-to-Machine (Ad hoc) 99

100 Scope and motivation Research question: How to evaluate the business opportunities of technical solutions and mechanisms that support the secondary use of spectrum allocated to a primary user? Project challenge: To connect the technical and business analysis 100

101 Agenda items Note on Methodology for cost & capacity analysis What can we observe if we look into wide area MBB access for capacity enhancement in urban areas What can we observe if we look into indoor MBB access services in offices and hot spot areas Summary 101

102 Agenda items Note on Methodology for cost & capacity analysis What can we observe if we look into wide area MBB access for capacity enhancement in urban areas What can we observe if we look into indoor MBB access services in offices and hot spot areas Summary 102

103 Agenda items Note on Methodology for cost & capacity analysis What can we observe if we look into wide area MBB access for capacity enhancement in urban areas What can we observe if we look into indoor MBB access services in offices and hot spot areas Summary 103

104 Analysis of business opportunities of secondary use of spectrum - The case of TV white space for mobile broadband access Jan Markendahl and Östen Mäkitalo Wireless@KTH, Royal Institute of Technology (KTH) Stockholm, Sweden

105 Spectrum as a resource Radio LMR TV GSM 0 GHz 1 GHz GSM 1 GHz 2 GHz UMTS ISM Radar LTE 2 GHz 3 GHz Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

106 Spectrum in 800 MHz band is re-allocated from TV to mobile broadband Radio LMR TV LTE GSM 0 GHz 1 GHz GSM 1 GHz 2 GHz UMTS ISM Radar LTE 2 GHz 3 GHz Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

107 What about allocated spectrum that is not used? Radio LMR TV LTE GSM 0 GHz 1 GHz GSM 1 GHz 2 GHz UMTS ISM Radar LTE 2 GHz 3 GHz Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,

108 What about allocated spectrum that is not used? Can we use it for other purposes - if we not cause interference to the primary user? 0 GHz Radio LMR TV LTE GSM 1 GHz GSM 1 GHz 2 GHz UMTS ISM Radar LTE 2 GHz 3 GHz Markendahl, Sanchez and Mölleryd COST-TERRA meeting, Brussels Nov 16,