4thGeneration Wireless Infrastructures: Scenarios and Research Challenges

1
4th-Generation Wireless Infrastructures
Scenarios and Research Challenges

• Aurelian Bria, Fredrik Gessler, Olav Queseth,
  Rickard Stridh, Matthias Unbehaun, Jiang Wu, and
  Jens Zander
• Royal Institute of Technology, KTH
• Maxime Flament, Chalmers University of Technology

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
2
Agenda

• Introduction
• Scenario-based Approach
• 3 Scenarios
• Working Assumptions
• Focal Areas of Research
• Conclusion
• Further Work

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
3
Introduction
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
4
Introduction Authors and purpose

• 4GW and PPC and their goals
• Fourth Generation Wireless Project, Personal
  Computing and Communications program
• Swedish effort researching future communication
  systems
• Paper shows
• How research was guided
• Results from workgroups

5
Introduction - What is 4G?

• Right now, we dont know - 4G mostly a buzzword
• Paper written in 2001, making predictions for
  2010
• Will 4G look anything like 3G?
• Following old trends, we expect
• Higher data rates
• New frequency bands for world wide standard
• However, situation not this straightforward since
  many communications systems are already in place

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
6
Introduction

• Why worry?
• Because from experience, need research now to
  deploy in a decade
• So how to identify relevant research topics?

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
7
Scenario-based approach
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
8
(No Transcript)
9
Scenario-based Approach

• Identifying reasonable assumptions
• Assumptions will change over span of study, as
  will external determinants on which assumptions
  are based

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
10
Scenario-based Approach

• "A scenario is a tool to explore a possible,
  plausible future by identifying the key technical
  and social developments required for it to be
  realized"
• Not predicting future predicting future
  possibilities to both prepare for and influence
  the future

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
11
3 Scenarios
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
12
Scenarios (1) Anything Goes

• Manufacturing companies dominate
• They advocate open de facto standards
• Use software to make flexible, multistandard
  equipment
• Many operators
• Fierce competition all over
• Affordability of services and equipment to most
  leads to narrowed social gap in society
• Consumer wins

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
13
Scenarios (2) Big Brother

• Privacy and security become major concern
• Need regulatory approval for everything
• Services, transport equipment provider reduced
  to a few trusted companies
• No competition, high regulation, high prices
• Consumer loses

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
14
Scenarios (3) Pocket Computing

• Not everyone can afford same service
• Different, specialized hardware and services
  targeting different groups
• Few operators and large manufacturers use
  standards to maintain dominance

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
15
Working Assumptions
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
16
Working Assumptions - Telepresence

• Virtual meeting
• Efficient data compression - 100 Mbps
• Most technically demanding

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
17
Working Assumptions - Information anywhere,
anytime

• Highly asymmetric traffic pattern
• Users with small portable terminals want large
  volumes of data, pictures and video

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
18
Working Assumptions - Interactive communication

• Example fridge telling repair shop its broken
• Computers already do this for updates
• Anything over 20 could have this

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
19
Working Assumptions - Security

• Will be integral
• Schemes that reliably prevent intrusions into
  private sphere will be in operation

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
20
Working Assumptions - Non-homogenous
infrastructure

• Everything digital
• Multitude of physical media/air interfaces
  inherited from earlier generations of wireless
• IP packet oriented switching
• Up to 100 Mbps for hand-portable use

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
21
Working Assumptions - Ad hoc unlicensed operation

• Spontaneous deployment in unlicensed 5 and 60 GHz
  bands competes with providers
• Technique for efficient multi-operator sharing of
  unlicensed spectrum developed
• Ad hoc structures provide part of infrastructure

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
22
Working Assumptions - Multimode access ports in
public systems

• Multiple access air interfaces to accommodate a
  wide range of terminals.
• Adaptive antennas that self configure to reduce
  cost.
• Access points in ad hoc to remain single mode

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
23
Working Assumptions - Terminals

• Terminals will have wide range of bandwidths
  10kbps 100Mbps.
• Battery life of at least one week.
• Terminals in the 5 and 60 GHz range to use
  adaptive antennas.

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
24
Working Assumptions - Conclusions

• Key problem is providing affordable high data
  rates everywhere.
• High data rates not surprising, challenge is
  making affordable
• Cost per bit transmitted constant, regardless of
  data rate
• Workgroups formulated to study working
  assumptions developed

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
25
Focal areas of 4GW Research
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
26
Research - Broadband OFDM Air Interface Design

• Working group assumption studied user-deployed
  access points and self-planning capabilities key
  factors in making 4GW infrastructure economically
  viable
• Short-range broadband wireless systems play key
  role in such a structure
• In most countries, this would involve the
  unlicensed 60GHz band
• 60GHz investigated indoors in office type
  environment, then shopping mall environment

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
27
Research - Broadband OFDM Air Interface Design
28
Research - Broadband OFDM Air Interface Design
29
Research - Broadband OFDM Air Interface Design

• High attenuation of 60GHz due to absorption by
  oxygen, therefore used to be frequency of choice
  for intelligence agencies
• Coverage not main limitation
• Unstable handover caused by fact that
  interference occurs in short bursts

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
30
Research - Broadband OFDM Air Interface Design

• Shown diversity on terminal side prerequisite for
  functioning system
• Examined effect of human body on shadowing
• Need site diversity to decrease probability of
  shadowing
• Claim it appears feasible to design wireless
  systems for high data rates that function in
  offices or public hot spots

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
31
Research - Smart Antenna

• Proposed to improve performance of short range
  systems
• 5GHz systems has greater range than 60GHz
• With dual arrays fulfills link capacity for 4GW
• Antenna array at user terminal since one
  wavelength 5mm
• Still need studies to compare QoS and coverage vs
  infrastructure cost

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
32
Research - Wireless Infrastructure architecture

• Can only achieve the required coverage and data
  rates if infrastructure costs are reduced by a
  few orders of magnitude.
• So what if wireless networks follow the WLAN
  paradigm of being deployed by customers?
• Would have to use unlicensed bandwidths, eg 17GHz
  and 60GHz.
• Dues to high free space loss, shadowing by
  humans, and attenuation by common building
  materials, need many APs

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
33
Research - Wireless Infrastructure architecture

• This group modeled office and shopping
  environment using these frequencies and found
  such environments tolerable to arbitrary
  placement of APs assuming they are reasonably
  uniformly distributed.
• In the shopping center case, some coarse planning
  is needed. 60 GHz needs too many APs.
• Outdoor scenarios not suited to user deployment
  approach

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
34
Research - Wireless Resource Management in
Multiple-Operator Infrastructures

• In future, either one operator in charge of each
  frequency band, or several that have to share.
• If they have to share, freq hopping appears to be
  best alternative.
• Also, isolation between infrastructures, smart
  antennas.

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
35
Research - Seamless IP mobility Support for
Mobile Applications

• Both wired and wireless networks for a single
  communication session
• Protocols must be flexible and robust due to
  unordered network infrastructure
• Authors believe that IP layer handover latency
  will be low enough to support realtime
  applications in the case of Mobile IP
• Basically internet and wireless integrating
  successfully over time

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
36
Conclusion
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
37
Conclusion

• Using scenario based approach, identified three
  possible scenarios
• Derived working assumptions from scenarios
• Found biggest obstacle to achieving goal to be
  constant cost per transmitted bit
• Therefore studying technologies that can break
  this cost/performance barrier

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
38
Conclusion

• Candidate technologies included
• User deployed infrastructure (ad hoc research)
• Advances in array signal processing

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
39
Conclusion

• Also, scenario based approach helped greatly in
  guiding thought process and setting up interfaces
  around subprojects to define each team's research
  responsibilities.

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
40
Further Work
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
41
Further Work

• Vast area, so group cannot be faulted for what
  they did not mention.
• Not difficult to pick other potential scenarios
• Paper talks about overcoming the spectrum
  problem, but never talks about alternatives
• For example, dynamic spectrum assignment
• Multimode terminals discussed, but not much is
  said about specific technologies to handle these
  issues
• For example, software radio

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
42
Thank you
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
43
Appendix
Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
44
Appendix

• What are 1G, 2G, 2.5G, 3G and 4G?Technically
  generations are defined
• 1G networks (NMT, C-Nets, AMPS, TACS) are
  considered to be the first analog cellular
  systems, which started early 1980s. There were
  radio telephone systems even before that.
• 2G networks (GSM, cdmaOne, DAMPS) are the
  first digital cellular systems launched early
  1990s.
• 2.5G networks (GPRS, cdma2000 1x) are the
  enhanced versions of 2G networks with data rates
  up to about 144kbit/s.
• 3G networks (UMTS FDD and TDD, cdma2000 1x
  EVDO, cdma2000 3x, TD-SCDMA, Arib WCDMA, EDGE,
  IMT-2000 DECT) are the latest cellular networks
  that have data rates 384kbit/s and more.
• 4G is mainly a marketing buzzword at the
  moment. Some basic 4G research is being done, but
  no frequencies have been allocated. The Forth
  Generation could be ready for implementation
  around 2012.

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith
45
Appendix

• Next-Generation Wireless Communications Concepts
  and Technologies
• Robert Berezdivin, Robert Breinig, and Randy
  Topp, Raytheon
• Propagation and Interference Issues in a 60 GHz
  Mobile Network
• Maxime Flament,
• Communication Systems group, Dept. of Signals and
  Systems

Naseem Hakim EE 360 Class Presentation Spring
2004 Professor Goldsmith