Broadband 2005 Current Status

1
Broadband 2005Current Status Future
Trends3G and BeyondJannie van Zyl 1st
November 2005
2
Agenda

• Vodacom Current Status
• GPRS
• EDGE
• UMTS
• Wireless Network Trends
• HSDPA / HSUPA
• WiMax
• OFDM
• UMA / Mesh Networks

3
Where are we Today?
4
First Mobile Phone ?
No Ringtones or SMS
5
First Mobile Car Phone ?
1924
1952
6
First Portable Mobile Phone ?
1973
7
New Terminal Capabilities
Multiplicity of Local Connectivity
Mobile Storage Revolution
New Ways of Displaying Content

• 20 GByte is enough for your entire music
  collection your entire photo album 20hrs of
  home movies 8 hours of DVD quality movies a
  bevy of games

8
New Form Factors
2007?
9
Where are we Today?

• Current South African Landscape

10
Vodacom Coverage gt 16M Subscribers

• 2G GPRS
• gt 6300 Base Stations
• gt 70 Land Area Coverage
• gt 97.5 Population Coverage
• gt 99.7 Availability
• 2G EDGE
• gt 37 Land Area Coverage
• gt 32 Population Coverage
• 3G WCDMA
• 1500 Node Bs (Base Stations)
• gt 99.3 Availability

11
EDGE Coverage Aug 2005
12
3G Coverage
13
Vodacom 3G Coverage

• Major Metros
• Holiday Areas
• All major airports
• Major hotels, conference venues and business
  centers
• Sites with high GPRS usage
• Coverage for key business partners
• Seamless Integration with 2G
• Target through-put of 300kbits/second avg.

14
Network Futures

• Overview

15
GSM Evolution towards 3G UMTS
2G
2G
3G
UMTS 1st Release
Theoretical maximum data speeds
GPRS
HSCSD
EDGE
GSM CSD
9.6 kb/s
57 kb/s
170 kb/s
470 kb/s
2048 kb/s
16
3G Evolution towards 4G
?
?
WiMax
70 Mb/s
4G OFDM MIMO
11 - 54 Mb/s
MBWA (Wide Area)
WiFi (Local Area)
50 - 100 Mb/s?
HSDPA/ HSUPA
15 Mb/s
UMTS (Dense)
14.4 Mb/s
UMTS (Wide Area)
OFDM Orthogonal FDM
2Mb/s
384 kb/s
Theoretical max data rates depend on radio
conditions /or system options eg bandwidth. Time
scales are approximate/illustrative. The terms
3G, Evolved 3G, Super 3G, Beyond 3G, 4G etc are
not officially defined.
3G 3G MBWA 4G
2003 2004 2006 2008
17
Network Futures

• Some Fundamentals

18
Constraints on Coverage

• Limit for a non-fading channel was established
  by Claude Shanon in a classic paper in 1948.
  Minimum received signal defined by

Eb / No -1.6dB
Energy per Bit
Noise Spectral Density

• Depends on
• A Fundamental Limit (Physics)
• Receiver Performance
• Interference

• Depends on
• Transmitter Power
• Size of Antennas
• Path Loss (Frequency)

• Modern systems are getting closer to this limit,
  so unless we use bigger antennas or more transmit
  power, more bits/second means smaller cells,
  whatever the radio access technology
• Uplink generally more constrained because of
  batter power and EMF safety limits

19
Constraints on Coverage

• coverage more sites

Range
20
Constraints on Coverage

• coverage more sites

• higher data rate more sites

3 Sector base station at 25m to outdoor PC card
Data Rate
Site density
100Mb/s 0.21km
50
10Mb/s 0.41km
15
3.6
1Mb/s 0.78km
100kb/s
1.49km
1
Range
21
Constraints on Coverage

• coverage more sites

• higher frequency more sites

• higher data rate more sites

3 Sector base station at 25m to outdoor PC card
Data Rate
Site density
100Mb/s 0.21km
50
10Mb/s 0.41km
15
3.6
1Mb/s 0.78km
100kb/s
1.49km
1
Range
22
Constraints on Coverage

• downlink efficiency (bps/Hz/site)

• coverage more sites

• higher frequency more sites

• higher data rate more sites

3 Sector base station at 25m to outdoor PC card
Data Rate
Site density
100Mb/s 0.21km
50
10Mb/s 0.41km
15
3.6
1Mb/s 0.78km
100kb/s
1.49km
1
Range
23
Constraints on Coverage

• coverage more sites

Fixed Mobile Chalk Cheese

• higher frequency more sites

• higher data rate more sites

3 Sector base station at 25m to outdoor PC card
Data Rate
Site density
100Mb/s 0.21km
50
10Mb/s 0.41km
15
3.6
1Mb/s 0.78km
100kb/s
1.49km
1
Range
24
Beware Specmanship My System is faster than
Yours

• Modern technologies (EDGE, HSDPA, WiMax, Flarion)
  ALL maximise use of the spectrum using flexible
  allocation of resources and adaptive modulation
  and coding.
• Downlink throughput often shared by all users
• What data rate will it deliver?
• How long is a piece of string?

25
Constraints on Coverage - Summary

• Modern technologies are beginning to approach
  Shannon limit - A new technology is unlikely to
  provide more than a marginal increase in range
  for a given data rate.
• The only way to get more coverage is to
• Transmit more power
• Transmit data slower
• Use bigger/higher antennas
• Use a lower frequency
• Get someone else to pass on a message
• Uplink transmit power likely to be the limiting
  factor in range
• Downlink transmit power and peak data rate
  determine downlink range. Peak data rate then
  needs to be shared between users

26
Mobility vs. Speed
User Data Rate
Broadband
Narrowband
High speed
M O B I L I Y

• Mobile
• Nomadic
• Fixed

Low speed
Level of mobility
Pedestrian
Nomadic
Indoor
Stationary
1
Mbps
0,1
10
Speed
27
Mobility vs. Speed

• Cellular technologies become more and more
  broadband
• (HSDPA, HSUPA)
• Alternative wireless technologies become
• more and more mobile
• (WiMAX, Flash-OFDM, 802.20)

• Several technologies compete for mobile
  Broadband
• HSDPA is the most promising candidate

28
Technology Landscape
NEXT Ad-hoc Mesh Networks
29
Wireless broadband the landscape
802.16-2004 (fixed WiMax)
802.16e (mobile WiMax)
Alvarion Navini Aperto Adaptix
Flarion / Qualcomm ArrayComm
Proprietary, but aspiring to WiMax certification
and eventually 802.16e
Mobile broadband
HSDPA
TD-CDMA/FDD IPWireless
Fixed wireless broadband
Mesh networks (mostly based on 802.11) SkyPilot Be
lAir Nortel WMN IWICS
30
Network Futures

• HSDPA / HSUPA

31
HSDPA Goals

• Efficient data traffic delivery mechanism that is
  fully compatible with the current 3GPP system
• HSDPA coverage is as good as UTRAN Rel99 No
  need for new sites
• Voice and data on the same carrier No need for
  extra spectrum to deploy
• Utilises current Network Infrastructure -
  Software upgrade
• Full mobility
• Improves spectral efficiency of the Rel99 system
  about 2-3 times for packet data services
• Significantly higher throughput and lower latency
  offers peak data rate of up to 14.4Mb/s (Practice
  1 2 Mb/s)
• Specification also supports data rate as high as
  384kbps in the uplink
• Requires new Terminals

When ???
32
HSUPA Goals

• Once again only a software upgrade on the Network
• Full mobility
• Improves spectral efficiency of the Rel99 system
  in the upload
• Significantly higher uplink throughput and lower
  latency
• Offers peak data rate of up to 5.76Mb/s, lt 100ms
• New Terminals
• Availability 2007?

33
Network Futures

• WiMax

34
Fixed WiMax Fixed Wireless Broadband

• Allocations available in the 3.5 5.8GHz bands
• Needs more cells than 3G
• high frequencies make migration to full mobile
  service problematic.
• Good solution for areas where DSL cant reach,
  but shear volume of data traffic makes it hard to
  compete for the heaviest internet households.
• Possible backhaul technology for WLAN public
  access points and pico-cells, but range and
  capacity restricts use as a base station backhaul
  solution.
• Intel remain bullish on its capabilities
• Intel CEO view better than DSL (50 100Mb/s).
• Our view peak rates lt10.5Mb/s - need MIMO or
  other techniques to improve it

Intel adds napa to Centrino roadmap
35
WiMax 802.16e (Mobile Wireless Access)

• There are no WiMax products available today
• Commercial products based on the superseded
  802.16a are not WIMAX certified
• These products are all focused on fixed wireless
  access deployments
• Site count required for mobile wireless is far
  greater than fixed
• Laws of physics have not changed, despite press
  speculation
• For 802.16e to offer comparable mobile service,
  similar densities as 3G needed
• Too early to know how 802.16e will perform.
• Simulations in a multi-cell deployment urgently
  needed
• Handover ??
• 802.16 includes variety of modes and intended
  applications
• FDD and TDD, licensed and unlicensed spectrum,
  massive number of options
• 70 Mbps capability and gt30 km range is true for
  only one mode not mobile!
• Historical focus on fixed wireless systems and
  wireless backhaul
• The only completed standard applies to fixed
  wireless access systems
• The 802.16e standard for mobility support is
  incomplete in many important areas

36
WiMax 802.16e (Mobile Wireless Access)

• Lack of detailed simulation or test results due
  to the standards not being finished
• Early results suggest that 802.16e may have a
  small increase in spectrum efficiency compared to
  HSDPA
• The exact performance figures for mobile WiMAX
  cannot be determined yet!
• Will only be possible once standard is complete
  and a WiMAX 802.16e profile has been selected
• Earliest 802.16e deployment likely to be 2007

37
Hype is Fading but Backers are Strong
2005 Strong Backers
2004 - hype
Wait and See

• Intels inclusion in Centrino makes WiMax mobile
  credible.

2005 set back
Competition in US from IPWireless?
38
Network Futures

• OFDM

39
Discontinuity in radio access technology
Tomorrow?
Today
Uplink evolution is not so clear!!
40
Flarions Flash-OFDM (Qualcomm)

• Overview
• Fully mobile technology with 7 years development
  effort
• Requires 2 x 1.25 MHz
• Successful Vodafone field trial
• Peak rates 2.7Mbps DL, 0.8 UL
• Aggregate 1.8Mbps DL, 0.4 UL
• lt40 ms round trip time (latency)
• Maturity
• No commercial deployments
• Infrastructure available from Flarion, Siemens
  and Nortel
• Flarion PCMCIA card

Excellent, but handover not as impressive..!
No service concept
41
Network Futures

• Various Considerations

42
Mobile TV - which technology?

• HSDPA effective for unicast of video clips and
  expansion of handset memory will allow non-real
  time viewing of off-peak download
• Serious adoption of 3G MobileTV will cause
  capacity problems then broadcast technologies
  will be needed

• MBMS is more cost effective than streaming
  LiveTV, but is limited to 4 concurrent TV
  channels
• We can expect to see mobile broadcast
  technologies rolled-out in Europe quite soon

Channel equivalents (per cell)

MBMS, 200kbit/ch
4
HSDPA, 200kbit/ch

3G, 200kbit/ch

1
5 users 3G capacitylimit
Concurrent viewers/cell
20 users 3G capacitylimit
Assuming a 3G carrier used only for MBMS
43
UMA (Unlicensed Mobile Access)

• Provides Access to GSM/GPRS services over
  alternative (potentially any) access methods
  current focus on WLAN targetted for Release 7.
• Concept has the potential to be extended to 3GSM
• Provides tight coupling and re-use of GSM/GPRS
  core network.
• Differs from existing 3GPP WLAN integration
  (I-WLAN) effort as it mainly benefits CS services
  (Voice, SMS).
• UMA primarily suited for extending CS into
  unlicensed spectrum
• I-WLAN primarily suited for extending broadband
  packet services into unlicensed spectrum
• BT utilising UMA for its Bluephone product.
• Launch has been delayed whilst they wait for
  commercial UMA solution.

44
Mesh Networks Cellular Topology

• Each node communicates only with closest base
  station.
• Base stations connected through high-capacity
  wires or optical fibers.
• Base stations do most of the work.
• Not a truly wireless network (only last hop is
  wireless) !

45
Mesh Networks Ad-Hoc Topology

• No base stations, but nodes can talk to each
  other.
• Nodes have to get organized without the help
  of base stations. This is a much more
  challenging problem!
• A truly wireless network!
• Another way to think about it a wireless
  internet.

46
Why Ad-Hoc Networks?

• They can be build very fast. No need to
  establish wired connections.
• They are very resilient. No single point of
  failure, such as a base station.
• They are spectrally more efficient than
  cellular networks. Every node can communicate
  with any other node, so nodes can make better use
  of the channel.
• Placing wires may be impossible, prohibitively
  expensive, or just not necessary.
• No need for operators.

47
Summary

• HSDPA / HSUPA provides best option for Mobile
  Wireless Broadband in the Short and Medium Term
• WiMax well suited to Fixed Wireless Broadband
• 802.16e not ready and is limited in Mobile
  Capability more Nomadic
• OFDM is the future for both Fixed and Mobile
  Networks 3 x W-CDMA
• Watch Qualcomm
• UMA and Ad-Hoc Networks holds great promise

48
Questions

• Jannie van Zyl
• jannie_at_vodacom.co.za
• Vodacom3G

49
Thank You