Spectrum Allocation

An Industry perspective

Mr Roberto Ercole
GSMA

© GSM Association 2009

 Issues

 Spectrum allocations for mobile

 Future developments in technology and the

implications for mobile broadband –
internationally harmonised spectrum and band
plans

 How this relates to the use of UHF in Asia

Pacific (698 to 806 MHz and the AWF/APT)

 The importance of the ITU and the need for an

IMT Agenda Item at WRC 16

© GSM Association 2009

 The need for mobile broadband
 This relates to the cost economics of coverage – a base station and low
cost devices is cheaper than laying fibre or upgrading copper local loop –
all other things being equal
 But for mobile to be that delivery platform requires that the cost
economics be as favourable as possible – to reach as many consumers
as possible at the lowest cost to consumers

wireless data growth in the World Bank

© GSM Association 2009
 Why Spectrum is important
 Need to ensure that there is enough spectrum to meet demand –
Capacity – various ways to meet capacity such as :
 Cell splits or Pico cells (higher cell density)
 Higher spectrum efficiency – GSM/EDGE to UMTS to HSPA
to LTE R8 to LTE R10
 More spectrum

 FCC paper Mobile Broadband: The Benefits of additional

spectrum - Oct 10 http://www.fcc.gov/Daily_Releases/Daily_Business/2010/db1027/DOC-
302324A1.pdf

 Benefits of extra spectrum in US of around $100bn over the next 5

years – assuming a baseline of 170 MHz . But 547 MHz available
in US can possibly be used under current rules.

 This 170 MHz vs 547 MHz highlights the impact of harmonisation

on device costs and usable spectrum


© GSM Association 2009
 Impact of Economies of Scale on device costs

Volumes mean initial 3G

WCDMA prices drop from
$600+ to around $100 now

© GSM Association 2009

 How is spectrum allocated between various services

 Three example countries in prime bands – 400 MHz

to 5 GHz:
– Malaysia

– UK

– US

© GSM Association 2009

 Spectrum use in Malaysia – 400 MHz to 5 GHz

Public
Government
Mobile

Radar 10.0% 14.5% Broadcasting

1.1%

19.3% 8.3% PM
MO
2.4% Bcast
SRD

Fixed Satellite
Links Satellite FL
PMSE
0.0%
24.9% RA
2.2% 15.9%
radar
Govt

14.5% to mobile, but this includes 2.6 GHz extension band, currently
used for BWA – without this only 10.2%

© GSM Association 2009

 UK Spectrum use 400 to 5000 MHz
Public mobile
With ext bands

Military
14.7%
24.7% 1.5%

8.0% PM
MO
2.4% Bcast
3.2% SRD
Fixed
Radar Links MSS
FL

14.2% PMSE
28.6% RA
radar
MOD

1.0% 1.6%
14.7% to mobile, but this includes 2.6 GHz extension band, currently not available
for mobile – without this only 10.3% - this 10.3% generates 50% of the economic
benefit of all spectrum use in the UK

© GSM Association 2009

 Spectrum use in the US – 400 to 5000 MHz
Military

Public
1.8% mobile
5.5%
Satellite 12.8%
(501 MHz)
7.0%
mil

10.9% PM
12.0% BC
MO
3.9%
PPDR

2.7% radar
MSS
2.2%
Radar 2.5% FG
sat
2.0%
aero
32.6%
SRD
NASA

Over 500 MHz identified for mobile – but most of the traffic on 170 MHz
spectrum according to FCC paper

© GSM Association 2009

 US National Broadband Plan identifies
future spectrum

© GSM Association 2009

 How does current technology add new bands

Current and foreseeable handset technology needs a

new RF front end per band . This need for separate
analogue components with high selectivity (Q) factor
means that the number of bands that can be
supported is limited.

© GSM Association 2009

 Why international harmonisation matters
 Because of the need for analogue components in the RF chain, effectively
means a new RF front end module per frequency band. But this is complex
and can impact on RF performance (as well as costs).
Radio Multiband Architecture
Example Number of bands a
MIMO
handset can support
LTE, HSPAevo,
FEM EDGE
Rx MIMO signal
is limited.
Diversity/MIMO front-end processing

As the number of

Baseband i/f
Main
FEM LTE, HSPAevo,
EDGE
Rx signal
bands goes up the
processing
RF efficiency goes
Main PA down.
‘Core’ WCDMA combination LTE, HSPAevo,
with international roaming EDGE
•Band 1, 2, 4, 5 (6), 8 Tx signal
•Quad band GSM processing
Add on
FEM
Additional band support:
e.g. band 3, 7, 11, 12-14, 20
or 40
Add on
FEM

Every new FEM adds to complexity

© GSM Association 2009
 Impact on quality of service

Static
Sensivity 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
dBm
-110
-109
-108
-107 X X
-106 X X X
-105 X X X
-104 X
-103 X
-102 X
-101 X
-100 X

Volume Based Performance Gains*

typically 1 dB per year
Introduction of smaller form Intro-
factor phones, dual band, duction
tri-band, direct conversion. of
GPRS

Sources: Ad Hoc measurements by RTT.

Industry sources 1992 to 2000.

The gain historically has been 1 dB per year : or 17% more sites for the same coverage !!
© GSM Association 2009
 How frequency bands are fragmenting
FDD
TDD
Band “Identifier” Frequencies (MHz)
Band “Identifier” Frequencies (MHz)
1 IMT Core Band 1920-1980/2110-2170
33,34 TDD 2000 1900-1920
2 PCS 1900 1850-1910/1930-1990 2010-2025
3 GSM 1800 1710-1785/1805-1880 35,36 TDD 1900 1850-1910
4 AWS (US & other) 1710-1755/2110-2155 1930-1990
5 850 824-849/869-894 37 PCS center gap (1915) 1910-1930
6 850 (Japan) 830-840/875-885 38 IMT extension center 2570-2620
gap
7 IMT Extension 2500-2570/2620-2690
39 China TDD 1880-1920
8 GSM 900 880-915/925-960
40 2.3 TDD 2300-2400
9 1700 (Japan) 1750-1785/1845-1880
10 3G Americas 1710-1770/2110-2170
Additional FDD and TDD
11 UMTS1500 1428-1453/1476-1501
3.5 GHz 3400-3600
12, US 700 698-716/728-746
3.7 GHz 3600-3800
13, US 700 776-788/746-758
14 US 700 788-798/758-768
Additional FDD
17 US 700 704-716/734-746
800 MHz 790-862

© GSM Association 2009

 Digital Dividend

 A practical example of the importance of frequency

harmonisation

© GSM Association 2009

 Effect of frequency on range and capex
Coverage of rural areas at about 30% of the cost of 2100 MHz

© GSM Association 2009

 UHF and Digital Dividend

 To ensure cost effective coverage in rural areas and will be required to boost
mobile broadband coverage
 The initial bands (72 MHz in Europe and 108 MHz in US) may not be enough and
more spectrum may be required in the longer term
– US national broadband plan has identified 120 MHz more UHF from
broadcasting
– Study by EU Commission suggested that extending the band to 698 MHz may
offer the best outcome for consumers and economic well being
 Figures vary but the impact on capex in rural areas can be around a 70% saving on
2 GHz core bands – 3dB link budget advantage

© GSM Association 2009

 AWF/APT and Digital Dividend – 2 x 45 MHz FDD

692 MHz* 703 MHz 748 MHz 803 MHz

Ext GB 5 45MHz Uplink 10 45 MHz Downlink 3 PPDR

698 MHz 758 MHz 806 MHz

Internal guard band

Centre Gap

Ext GB external guard band with broadcasting

*692 MHz due to China – PRC
Broadcasting plan

Source AWF9-/OUT-13 – Seoul Sept. 10 (also includes a TDD option)

© GSM Association 2009
 AWF process importance to UHF band plan

 With support from Asia-Pacific (over half world’s population) it has

enough scale to drive down costs
 The GSMA understands that China has a preference for the
inclusion of a TDD band
 In general the industry has a preference for FDD in this band (as it
has a preference for TDD in 2.3/2.4 GHz band)
 Indian submission to AWF has noted the benefit of FDD in this band
for improved rural coverage (AWF-7/INP-47 (Rev.1)).
 Support for the AWF band UHF plan/s will boost the chance of
success in achieving the necessary economies of scale.
 Without support from R3 Administrations ,the danger is fragmentation
and that only US or EU (FDD) equipment will be available.
 An AWF FDD band plan may advantage R3 manufacturers

© GSM Association 2009

 WRC12 and WRC16 – GSMA objectives

 New Agenda item for approval at WRC12 to be

discussed at WRC16

© GSM Association 2009

 WRC and more IMT Spectrum
 GSMA proposes that an Agenda item for WRC16 on IMT spectrum

 Specifically for WRC12 Agenda Item 8.2, that for the WRC 16
agenda : “To consider the frequency bands identified for IMT
with a view to rationalising, consolidating, and expanding these
as appropriate, with the objective of achieving internationally
harmonised bands, preferably on a global basis.”

 GSMA has submitted this to CEPT (PTA) and is developing a draft

resolution, and will submit a proposal to APG in December as well as
to other regional groups, and has been proposed in Citel

 GSMA and industry generally believes that WRC12 needs to ensure

that there is an Agenda item for WRC to consider the issue

© GSM Association 2009

 Why the industry believes this is needed

 Mismatch between bands in some regions (R1 UHF 790-862 MHz

and R2/R3)
 Complex structure of footnotes and rules
 More spectrum may well be required for mobile in the time frame of
the year 2020 (when WRC16 identification may become available
nationally)
 The industry is undertaking studies to look at demand and supply
factors on spectrum (technology, regulation, and demand curves)
 GSMA believes that a number of R3 countries might support such a
proposed Agenda Item.

© GSM Association 2009

 Conclusion

 Spectrum is important factor in costs of supplying mobile broadband

 Not only the amount of spectrum, but where it is in the band and the number of
countries that will support it
 Strong signals from Administrations (WRC/AWF etc) are required to provide the
stimulus for industry to commit to development costs
 The current and likely future design of RF kit (especially handsets) will limit the
number of bands and air interfaces that can be supported
 This band aggregation issue, means that bands need to “compete” for space in
handsets

© GSM Association 2009