1. Department of Systems & Computer Engineering
CELLULAR EVOLUTION
Halim Yanikomeroglu
Department of Systems & Computer Engineering
Carleton University
Ottawa, Canada

2. Cellular Basics Importance of standards
Tedious standardization process, amortization period  delay
Generations of technologies: 1G, 2G, 3G, 4G, 5G
Confusing terminology
Role of ITU (circular letters)

3. Cellular: Earlier Generations

4. Cellular Generations – A More Detailed Look
1G: AMPS, G: GSM, 1991

5. Cellular Generations – A More Detailed Look
1G: AMPS, G: GSM, 1991
3GPP Platform: “unites 6 telecom standard development organizations (ARIB, ATIS, CCSA, ETSI, TTA, TTC), and provides their members with a stable environment to produce the highly successful Reports and Specifications that define 3GPP technologies”. (Other platforms and organizations: 3GPP2, IEEE, …)

6. Cellular Generations – A More Detailed Look
1G: AMPS, G: GSM, 1991
R99 R4 R5 R6 R7 R8 R9
R10
R11
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
3GPP Platform

7. Cellular Generations – A More Detailed Look
1G: AMPS, G: GSM, 1991
R99 R4 R5 R6 R7 R8 R9
R10
R11
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
LTE
LTE
Adv
HSPA+
UMTS
HSPA DL
HSPA UL
3GPP Platform

8. Cellular Generations – A More Detailed Look
1G: AMPS, G: GSM, 1991
ITU-R IMT-2000 circular letter ITU-R IMT-Advanced circular letter
R99 R4 R5 R6 R7 R8 R9
R10
R11
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
LTE
LTE
Adv
HSPA+
UMTS
HSPA DL
HSPA UL
3GPP Platform

9. Cellular Generations – A More Detailed Look
1G: AMPS, G: GSM, 1991
ITU-R IMT-2000 circular letter ITU-R IMT-Advanced circular letter
3G: IMT-2000 compliant G: IMT-Advanced compliant
R99 R4 R5 R6 R7 R8 R9
R10
R11
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
LTE
LTE
Adv
UMTS
HSPA DL
HSPA UL
HSPA+
3GPP Platform

10. Cellular Generations – A More Detailed Look
1G: AMPS, G: GSM, 1991
ITU-R IMT-2000 circular letter ITU-R IMT-Advanced circular letter
3G: IMT-2000 compliant G: IMT-Advanced compliant
R99 R4 R5 R6 R7 R8 R9
R10
R11
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
LTE
LTE
Adv
UMTS
HSPA DL
HSPA UL
HSPA+
3GPP Platform
Release 12 Time Plan:
Stage 1 freeze – Mar 2013
Stage 2 freeze – Dec 2013
Stage 3 freeze – Jun 2014

11. Cellular Generations – HSPA and LTE Users
4gamericas.org

12. Cellular Connections – 2G, 3G, 4G

13. Cellular Generations
DATA

14. Mobile device for everyone
Cellular Generations 4G
Mobile device for everyone 3G
Gbps
data 2G
Mbps
Mbps
data 1G
kbps
kbps
2010 – will be the decade of LTE
2020 -> - will be 5th G what will it hold?
Not entirely sure but – what is sure is that Between Now and then-
Cell size will definitely shrink – and consequently
Cell count will increase
All in an order of magnitude
Devices will undoubtedly expand exponentially
Data M2M will explode – And this will inevitably lead to a concatenation of growth trends
The growing need for information and interaction will spur a number of Mega-trends ……
data
AMPS
bps
bps
AMPS
Source: Huawei (circa 2010)
1980
1990
2000
2010
Time
2020 14

15. Mobile device for everyone
Cellular Generations 5G 4G ? ?
data
Mobile device for everyone 3G
Gbps
data 2G
Mbps
Mbps
data 1G
kbps
kbps
2010 – will be the decade of LTE
2020 -> - will be 5th G what will it hold?
Not entirely sure but – what is sure is that Between Now and then-
Cell size will definitely shrink – and consequently
Cell count will increase
All in an order of magnitude
Devices will undoubtedly expand exponentially
Data M2M will explode – And this will inevitably lead to a concatenation of growth trends
The growing need for information and interaction will spur a number of Mega-trends ……
data
AMPS
bps
bps
AMPS
Source: Huawei (circa 2010)
1980
1990
2000
2010
Time
2020 15

16. Mobile device for everyone
Cellular Generations 5G 4G ? ?
data
Mobile device for everyone 3G
Gbps
data 2G
Mbps
Mbps
data 1G
kbps
kbps
2010 – will be the decade of LTE
2020 -> - will be 5th G what will it hold?
Not entirely sure but – what is sure is that Between Now and then-
Cell size will definitely shrink – and consequently
Cell count will increase
All in an order of magnitude
Devices will undoubtedly expand exponentially
Data M2M will explode – And this will inevitably lead to a concatenation of growth trends
The growing need for information and interaction will spur a number of Mega-trends ……
data
AMPS
bps
bps
AMPS
Source: Huawei (circa 2010)
1980
1990
2000
2010
Time
2020 16

17. Direction 1: Highly Capable Terminals

18. Direction 1: Highly Capable Terminals

19. Direction 2: IoT – Integration of Physical and Digital Worlds

20. Direction 2: IoT – Integration of Physical and Digital Worlds
ITU-T Rec. Y.2221 ( Requirements for support of Ubiquitous Sensor Network* applications and services.
* This is the ITU-T term for MTC systems and applications involving possible large numbers of sensors potentially spread over a large area
Sensor Networks monitor physical or environmental conditions, (e.g., temperature, sound, vibration, pressure, motion or pollutants) at various locations. These sensors need to communicate with a server.
Early sensor networks operate in isolation. Networked sensor applications enable new possibilities for consumers, public organizations, enterprises, government, etc.
Applications integrate data from sensor networks to achieve industrial or home automation control, agricultural monitoring, healthcare, environment and pollution monitoring, and disaster surveillance, security, etc.

21. Access to Information
Books
Brick library 21

22. Access to Information
Books
Brick library
Internet library 22

23. Access to Information Books Brick library Internet library
Data of all sorts 23

24. Access to Information Books Brick library Internet library
Data of all sorts
Easy access (I/F) 24

25. Access to Information Books Brick library Internet library
Data of all sorts
Easy access (I/F)
Available before you ask/think 25

26. Diversification of Applications and Scenarios
Old Cellular:
unimodal  Optimized for one application in one scenario
Voice, outdoor, high power, mobile

27. Diversification of Applications and Scenarios
Old Cellular:
unimodal  Optimized for one application in one scenario
Voice, outdoor, high power, mobile
New Wireless:
multimodal  Will have to be optimized for various applications in various scenarios
Voice, video, haptics, 3D, …
Outdoors, indoors
Centralized, distributed/autonomous
Scheduled, contention-based
Human operated, MTC
Rate – delay – reliability – energy …

28. 5G Requirements x10 – x20: Peak rates x100 – x1000: Area rates
x10 – x100: Device density
x0.1: Latency
x10 – x100: Energy efficiency

29. 5G Requirements x10 – x20: Peak rates x100 – x1000: Area rates
x10 – x100: Device density
x0.1: Latency
x10 – x100: Energy efficiency
LTE-A is already a very ambitious standard
Some of the above will have to wait for 6G!

30. Latency in 3G & 4G Networks30

31. Evolving Performance Metrics
Bits/sec/Hz 31

32. Evolving Performance Metrics
Bits/sec/Hz
Bits/sec/Hz/km2 32

33. Evolving Performance Metrics
Bits/sec/Hz
Bits/sec/Hz/km2
Bits/sec/Hz/km2/$ 33

34. Evolving Performance Metrics
Bits/sec/Hz
Bits/sec/Hz/km2
Bits/sec/Hz/km2/$
Bits/sec/Hz/km2/$/joule 34

35. Time for 5G Research?
4G: 3GPP rel-8 (LTE), rel-9, rel-10 (LTE-A), rel-11, rel-12 (?), rel-13 (?)
5G: not defined at this point
(keep an eye on EU framework program 8 projects: 2014 – 2020)
Research
Standardization
Deployment 5G ? 4G
LTE, LTE-A, m 3G
1X EV-DO, HSPA, HSPA+
10s-100s Gbps ?
100 Mbps mobile
1 Gbps nomadic
(IMT-Advanced compliant)
Mbps

36. Where We Are Now and Where We Are Heading To
R8, R9, R10 LTE/LTE-A

37. Key Technologies for LTE/LTE-Advanced (R8, R9, R10)
OFDM
MIMO
Spectrum aggregation
HetNet, Relay [to be matured]
CoMP (coordinated multipoint) [moved to R11]
 A number of LTE/LTE-A technologies are ahead of their time

38. Where We Are Now and Where We Are Heading To
R8, R9, R10 LTE/LTE-A
R11, R12 (Mar 2013, Dec 2013, Jun 2014), R13

39. 3GPP Release 12 Workshop Ljubljana, 11-12 June 2012
Priority Areas
Higher data rates
More capacity
Complimentary Areas
Energy saving
Cost efficiency
Support for diverse application and traffic types
Backhaul enhancements
Incredible resource
State-of-the-art in 3GPP

40. 3GPP Timelines
Panasonic
Samsung

41. Some Key Technologies for beyond LTE-A
Spectrum aggregation
MIMO (multi-layer, adaptive beamforming)
Multihop relaying
Terminal relaying (cellular-assisted ad hoc)
Advanced CoMP (cloud-RAN)
HetNet (heterogeneous networks)
SON (self-organizing, self-configuring, self-healing networks)
FeICIC (further enhanced intercell interference coordination)
Interference cancellation
MUD (multiuser detection)

42. HetNet (Heterogeneous Network) Architecture
Across network routing

43. Where We Are Now and Where We Are Heading To
R8, R9, R10 LTE/LTE-A
R11, R12 (Mar 2013, Dec 2013, Jun 2014), R13
EU Framework Program 8, Horizon 2020 (2014 – 2020)

44. Where We Are Now and Where We Are Heading To
R8, R9, R10 LTE/LTE-A
R11, R12 (Mar 2013, Dec 2013, Jun 2014), R13
EU Framework Program 8, Horizon 2020 (2014 – 2020)
5G PPP – The 5G Infrastructure Public Private Partnership (2014)

45. Where We Are Now and Where We Are Heading To
R8, R9, R10 LTE/LTE-A
R11, R12 (Mar 2013, Dec 2013, Jun 2014), R13
EU Framework Program 8, Horizon 2020 (2014 – 2020)
5G PPP – The 5G Infrastructure Public Private Partnership (2014)
ITU WRC 2015

46. Where We Are Now and Where We Are Heading To
R8, R9, R10 LTE/LTE-A
R11, R12 (Mar 2013, Dec 2013, Jun 2014), R13
EU Framework Program 8, Horizon 2020 (2014 – 2020)
5G PPP – The 5G Infrastructure Public Private Partnership (2014)
ITU WRC 2015
ITU circular letter: IMT-2020 5G

47. Where We Are Now and Where We Are Heading To
R8, R9, R10 LTE/LTE-A
R11, R12 (Mar 2013, Dec 2013, Jun 2014), R13
EU Framework Program 8, Horizon 2020 (2014 – 2020)
5G PPP – The 5G Infrastructure Public Private Partnership (2014)
ITU WRC 2015
ITU circular letter: IMT-2020 5G
Beyond…

48. Time Scales Near-term: Towards 2020 (4G evolution)
Middle-term: Around (5G)
Long-term: Beyond (5G evolution) 48

49. Time Scales Near-term: Towards 2020 (4G evolution)
Middle-term: Around (5G)  around the corner
Long-term: Beyond (5G evolution) 49

50. Time Scales
1980
1985
1990
1995
2000
2005
2010 1G 2G
2.5G 3G
3.5G 4G
voice

51. Time Scales
1980
1985
1990
1995
2000
2005
2010
2015
2020 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
voice
data (video)

52. Time Scales
1980
1985
1990
1995
2000
2005
2010
2015
2020 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
voice
data (video)
IoE

53. Time Scales IoE concepts are ahead of technology 1980 1985 1990 1995
2000
2005
2010
2015
2020 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
voice
data (video)
IoE

54. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
voice
data (video)
IoE

55. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice
data (video)
IoE

56. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice L1

57. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice
data (video) L1 L2

58. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice
data (video) L1 L2 L3

59. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice
data (video)
IoE L1 L2 L3 VN
cloud

60. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice
data (video)
IoE L1 L2 L3 VN
cloud
computing

61. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice
data (video)
IoE L1 L2 L3
optimization VN
cloud
math
computing

62. Time Scales IoE concepts are ahead of technology no time! 1980 1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice
data (video)
IoE L1 L2 L3
differential
topology
number theory VN
cloud
math
machine learning AI
computing

63. more interdisciplinary
Time Scales
IoE concepts
are ahead
of technology
no time!
1980
1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
2035 1G 2G
2.5G 3G
3.5G 4G
4.5G 5G
5.5G 6G
voice
data (video)
IoE L1 L2 L3
differential
topology
number theory
more people
more resources
more interdisciplinary
more collaborations VN
cloud
math
machine learning AI
computing

64. ICT: Mother of All Sectors…
Energy
Health
Entertainment
Transportation
Automotive
Agriculture
5GICT
Defence
Public Safety
Education
Hospitality
Municipalities

65. Resource Block (RB) Courtesy of Jing Dang RE：Resource Element
Frequency domain: 15 kHz (one subcarrier)
Time domain: one OFDM symbol (1/14 ms)
frequency
REG：RE group, REG = 4 RE
One subcarrier
CCE：Control Channel Element, CCE = 9 REG
time
One OFDM
symbol
RB：Resource Block
RB = 84 RE
This figure shows one RB:
7 OFDM symbols in time domain (0.5 ms, one slot)
12 subcarriers in frequency domain (180 KHz)
Courtesy of Jing Dang

66. LTE-TDD Frame Structure
LTE-TDD DL/UL configuration
One Frame
Time duration: 10 ms
Two half frame (5 ms each)
10 subframes (1ms each)
Two slots per subframe (0.5 ms each)
DL-UL Configuration
Switch-point
periodicity
Subframe number 1 2 3 4 5 6 7 8 9
5 ms D S U
10 ms
Courtesy of Jing Dang

67. Reuse and Interference
Channel reuse
Co-channel interference, multiple access interference
Radio access network (RAN)
Denser frequency reuse
Increased capacity
Increased interference
Decreased quality

68. Resource Reuse Schemes
Soft frequency resue
Partial frequency reuse

69. Intercell Interference Coordination (ICIC)
Reuse factor: 1 / cluster size 1G, 2G: 1/7, 1/4 3G: 1/3 4G:  1 Ultimate reuse factor: 1 per cell (sector) Conventional static (a priori) resource allocation (scheduling): For the entire leased spectrum, or a big portion of it One reuse factor ICIC: Dynamic (aware) resource allocation for each RB, taking the channel and traffic into account Different reuse factor for each RB

70. Dynamic Design
Static design: Can not cope up with channel and traffic variations
Static and a priori resource allocation  Dynamic resource allocation

71. Dynamic Design
Static design: Can not cope up with channel and traffic variations
Static and a priori resource allocation  Dynamic resource allocation
ICIC: Intercell interference coordination (R8 – LTE)
eICIC: enhanced ICIC (R10 – LTE-A)
FeICIC: Further enhanced ICIC (R11, R12)

72. 2G Limited cooperation between APs (for handoff)
No cooperation between UEs
Interference: handle with fixed assignments  not a great concern
RRM: easy; circuit-switched CBR applications  power control
Perfect each AP-UE link  PHY

73. 3G/3G+/4G- Limited cooperation between APs No cooperation between UEs
Smaller cells
Denser reuse (every cell, every sector)
Interference: concern
Fractional Frequency Reuse (FFR)
Soft Frequency Reuse (SFR)
Scheduling: important

74. 4G HetNets (femto-/pico-APs, relay)
Cooperation between APs (ICIC, eICIC)
No cooperation between UEs
Scheduling: very important
Interference: may become unpredictable, becoming a concern

75. 4G+/5G
Hi-HetNet (C-RAN, femto-/pico-APs, DAS, various types of relays including terminal relays)
Intense cooperation between select APs (feICIC, CoMP)
Cooperation between UEs
Interference: highly unpredictable (due to autonomous RRM decisions); major concern
 sophisticated, robust, good (not necessarily optimal) decisions
 partially centralized, partially distributed (opportunistically)
 learning (artificial intelligence)

76. 5G+ Indoors: # of APs >> # of UEs
Short distance, dedicated links
Optimized air interface
60-90 GHz carrier, FSO
Highly directional antennas
Super ultra rates
Atto-cell + FTTDesk
Outdoor hot-spots: # of APs << # of UEs
Mesh connectivity
Issues similar to previous slide

77. Evolution of Networks 1G 2G N=7 N=3 N: Cluster size ↓
interference is a concern 77

78. Evolution of Networks 1G 2G 3G N=7 N=3 FFR N: Cluster size ↓
interference is a concern 78

79. Evolution of Networks 1G 2G 3G 4G N=7 N=3 FFR ICIC, eICIC
N: Cluster size ↓ HetNet
interference is a concern 79

80. Evolution of Networks interference is a concern
centralized – CRAN, VRAN
1G G G G G
N=7 N=3 FFR ICIC, eICIC
N: Cluster size ↓ HetNet HiHetNet
interference is a concern 80

81. interference is a concern
Evolution of Networks
centralized – CRAN, VRAN
1G G G G G
N=7 N=3 FFR ICIC, eICIC distributed/autonomous
N: Cluster size ↓ HetNet HiHetNet random access
interference is a concern 81

82. Evolution of Networks centralized – CRAN, VRAN 1G 2G 3G 4G 5G
N=7 N=3 FFR ICIC, eICIC distributed/autonomous
N: Cluster size ↓ HetNet HiHetNet random access
Atto-cell
interference is not a concern 82

83. Small Cell Deployment

84. Small Cell Deployment
Interference ↑

85. Stochastic Geometry
Source: U of Texas, Austin

86. Traffic Generation Maximum homogeneity: Lattice
Sub-Poisson: perturbation
Complete-randomness: Poisson
Sub-Poisson
Poisson
Super-Poisson
Super-Poisson:
Time domain: MMPP, HMM, HHMM (NHMM)
Space domain:
Clustering Perturbation
Courtesy of Meisam Mirahsan
and Dr. Rainer Schoenen

87. Advanced RAN with Advanced RRM
Any fixed assignment is inefficient
cannot adapt to or exploit channel and traffic conditions
All decisions are dynamic and opportunistic
No a-priori partitioning of radio resources
No WT-BS assignment (dynamic routing in the mesh)
Reuse may be > 1
Wired elements (BS, DA) and fixed relays:
Cooperative RRM for interference management and avoidance
Nomadic, moving, and terminal relays:
Robust, distributed, plug-and-play, low-overhead, sub-optimum RRM algorithms
 cognitive radio (spectrum, OSA), dynamic feedback control, machine learning, artificial intelligence  inter-disciplinary
Very different from conventional cellular networks

88. Shift in Emphasis
A-RAN: Advanced radio access network  HetNet, Cloud-RAN
A-RRM: Advanced radio resource management (layer-2 & -3)
 cooperation, coordination, collaboration
A-PHY: Advanced physical layer
Well-integrated advanced RRM and advanced PHY in the presence of a powerful RAN, for
handling interference
handling non-uniform traffic
maximization of the utilities
Cross-layer and across-network cooperation/coordination/collaboration
link  cell  network (not cellular in the classical sense)

89. Emerging Tools Optimization, Stochastic Optimization
↓ Stochastic Geometry
Game Theory ↓
Machine Learning, Control Theory
Artificial Intelligence
Ex: Cognitive radio: sense, decide, learn