The document discusses research on enabling sharing and efficiency in wireless networks beyond LTE. It discusses how sharing can improve coverage through infrastructure sharing between operators and capacity through techniques like multi-user MIMO across small cells. Current areas of investigation include quantifying efficiency gains from increased resource sharing based on correlation in demand, incentive mechanisms for sharing, and virtualization to enable networks without borders composed from shared resources. The goal is to develop mechanisms and economic models to efficiently share resources across technologies and ownership models.

1. LTE and Beyond in a Sharing Economy
Luiz DaSilva
Stokes Professor in Telecommunications, Trinity College Dublin
Professor, ECE, Virginia Tech
… with Danny Finn, Paolo Di Francesco, and Jacek Kibiłda
I International Workshop on Challenges and Trends for Broadband Mobile
Networks – Beyond LTE-A
Campinas, Brazil, 6 November 2013

2. Beyond LTE-A
Networks without Borders
Sharing and macro-cells: efficiencies in
coverage and capacity
Sharing and small cells: MU-MIMO
The road ahead

3. Beyond LTE-A
• Smartphones use 24x
more data than regular
phones
• Tablets use 122x more
data than smartphones
• However, it is not feasible
for operators to increase
prices proportionally to
this demand

4. • More bands
licensed for LTE
• TVWS
• 3.5 GHz (US)
• 2.3 GHz (UK)
• 30 GHz
• Pico/femto cells
• Operator-deployed
WiFi
• Ultra-dense cells in
unlicensed spectrum
(60 GHz)
SHARING
• Interference
tolerance
• Massive MIMO
• CoMP

5. Core research
fundamental principles that will allow the wireless network of
the future to evolve into new architectures characterized by
increasing autonomy, resource sharing, and ubiquity of
wireless services
ability to learn
distributed and
autonomous decision
making
transient ownership of
resources

6. Networks without Borders
• Network composed on the run from a pool of resources
(spectrum, infrastructure, management services, …)
• Contributors to this pool range from households to small scale
operators to traditional wireless providers
• Network exists, virtually, to provide specific services to a specific
subscriber/user population
• Virtualization is a key component, leading to new entities (the
resource aggregator, the virtual architect)
• New business models

7. An evolution…
Infrastructure
Infrastructure
Infrastructure
Service Provider
Service Provider
Service Provider
Mobile Provider A
Mobile Provider B
Mobile Provider C

8. An evolution…
Infrastructure
Infrastructure
Infrastructure
Service Provider
Mobile Provider A
Service Provider
Mobile Provider B
Service Provider
Mobile Provider B

9. An evolution…
Infrastructure
Infrastructure
Service Provider
Mobile Provider A
Service Provider
Mobile Provider B
Service Provider
MVNO C

10. Networks without Borders
Infrastructure
Wholesaler
Resource
Aggregator
SERVICE
(Call, Game,
Content, ....)
Service Provider
Virtual Architect
Traditional
Mobile
Operator
Business or
Enterprise
Infrastructure
Provider
Household
Infrastructure
Provider
Individual
Infrastructure
Provider
Spectrum
Provider
Resource Pool
L. DaSilva, J. Kibilda, T. Forde, P. di Francesco, and L. Doyle, “Customized Services over Virtual Wireless
Networks: The Path towards Networks without Borders,” Proc. FMNS, July 2013

11. Increased efficiency and lower costs through…
❶ Incentives for the deployment of localized (small cell,
primarily) infrastructure by medium-sized and small operators
❷ The ability to provide service over infra-structure that
employs heterogeneous technologies, and has different
properties and ownership
❸ Improved service in currently under-served areas
❹ The ability to offer virtual wireless networks with different
associated quality of experience, at different price points

12. Current areas of investigation…
•
Mechanisms and APIs to enable
aggregation of resources
•
Pricing and market models for a fully
virtualized wireless network
•
Incentives for continued investment in
infrastructure
•
Public interest rationales for regulation, to
ensure competitive pricing and service level
outcomes
L. E. Doyle, J. Kibilda, T. K. Forde, and L. A. DaSilva, “Spectrum without Bounds, Networks without
Borders,” Proceedings of the IEEE, 2014 (submitted)

13. Sharing and macro-cells: coverage and capacity
Network shaping: Architect a network that meets the service
requirements at a minimum resource cost

14. Coverage optimization
min
*𝑥 𝑗 ,𝑧 𝑖 +
s.t.:
𝑐𝑗 𝑥𝑗
𝑗∈S
a 𝑖𝑗 𝑥 𝑗 ≥ 𝑧 𝑖 , ∀𝑖 ∈ 𝑃
𝑗∈𝑆
1 − 𝑧 𝑖 log(Pr 𝜉 𝑖 = 0 ) ≥ log(𝑝)
𝑖∈𝑃
𝑥 𝑗 ∈ 0,1 , ∀𝑗 ∈ 𝑆
𝑧 𝑖 ∈ 0,1 , ∀𝑖 ∈ 𝑃
𝑟 𝑖𝑗
Where 𝑎 𝑖𝑗 = 𝕀( 𝑟 ∗ ≥ 1), 𝑝 denotes pre-specified reliability level and 𝜉 𝑖 denotes service
𝑖
request coming from pixel 𝑖

15. GSM900 BS
GSM1800 BS
UMTS BS
Intra-operator
Co-located [%]
Inter-operator
co-located [%]
Warszawa
514
423
174
337
54.9
4.5
Wrocław
273
207
122
229
66.3
8.1
Olsztyn
74
56
37
68
79.7
5.4
Świdnica
29
27
13
20
65.5
6.9
Area
BS
Case study

16. Coverage sharing – efficiency results
Efficiency gain through infrastructure sharing for uniform deployment and Polish
case study; a) homogeneous power allocation, b) heterogeneous power allocation
J. Kibiłda and L. DaSilva, “Efficient Coverage through Inter-operator Infrastructure Sharing in Mobile
Networks,” in Proc. of Wireless Days, November 2013.

17. Traffic dynamics
• Dataset from Irish operator
(Meteor)
• Data sessions (2G/3G)
• Voice call records (2G/3G)
• More than 10.000 transmitters
to be analyzed
• Better understanding of traffic dynamics in cellular networks
• Temporal characteristics
• Spatial characteristics
• Spatio-Temporal characteristics
• Correlation in demand
• Assess correlation in demand combining datasets from different
operators (e.g. Meteor and O2) and other publicly available data (e.g.
demographic data on population density)

18. Correlation in time
• Hourly usage
• Clear daily trends – high peaks
at 24h interval and low peaks
12h offset
• The aggregated network traffic
shows good temporal
correlation
• Individual base stations do not
show the same good
correlations, but they keep the
periodicy
Autocorrelation – Meteor data

19. Correlation in space
• Hourly usage
• The periodicity has
disappeared
I
N
 w
i
j
 w

i
ij
j
ij
xi  x x j  x
i
xi  x
• Binary weight coefficient (ij)

to consider only base stations
that are relatively close
(distance < 2.5 km)
• Overall correlation is small
when considering the network
as a whole
• Smaller areas (e.g. Dublin)
show higher correlation, but still
relatively low
Morans I statistic –
Meteor data

20. Current areas of investigation…
•
Quantifying the expected efficiency gain
from increased resource sharing and its
relationship to correlation in demand
experienced by infrastructure providers
•
Stochastic models of infrastructure
deployment and study of the impact of
different infrastructure density and
distribution on the potential efficiency
gains from sharing
•
Game theoretic models of incentives and
preferences from the different players in
this architecture, capturing the geographic
nature of wireless access resources

21. Sharing and small cells: MU-MIMO
Examples of sharing in small cells:
• Open subscriber groups
• Small cells as infrastructure contributors to virtual
networks (Networks without Borders)
• Small cells operating in shared spectrum (e.g., 3.5
GHz)
• MU-MIMO across small cells

22. MU-MIMO
UE1
eNB
UE2
• With MU-MIMO, multiple UEs are spatially multiplexed on
different beams within the same time/frequency resource block
• Co-scheduled users must have orthogonal precoders

23. In small cell scenarios…
•
Fewer UEs per cell
• Fewer UEs for MU-MIMO pairing
10 m
10 m
10 m
10 m
10 m
•
Denser deployment so more cells within range
• If we reassign UEs between neighbouring cells, can we
increase UE throughputs by creating additional MUMIMO pairs?

24. Simulation and results
•
System level simulation for different
small cell and user densities
•
3GPP Dual Stripe small cell scenario
•
We found that:
4% of UEs get reassigned
9.6% gain to reassigned UEs
13.1% gain to target UEs
(DR = 0.2, 𝑆𝐸 𝑎𝑣𝑒 )
• To put this in perspective:
35% increase in MU-MIMO gains from
in 𝑆𝐸 𝑎𝑣𝑒 case
(with 4UEs/small cell, DR = 0.2)
… and even higher increases with
fewer UEs/small cell

25. Current areas of investigation…
•
Refinements to the mechanism based on
different UE scheduling methods
•
Extension for energy efficiency applications
• MU-MIMO-based UE reassignment
used to intelligently create empty cells
which can be temporarily switched off
•
Consideration of outdoor scenarios
D. Finn, H. Ahmadi, A. Cattoni, and L. A. DaSilva, “Multi-User MIMO across Small Cells,” IEEE ICC, 2014
(submitted)

26. The Road Ahead
•
“The new status symbol isn’t what
you own – it’s what you’re smart
enough not to own” – Lynn Jurich
•
We want to develop the resource
management mechanisms,
interfaces, and economic models to
enable sharing across technologies
and ownership models
•
We are evaluating our resource
management approaches using real
wireless deployment and usage
data

27. Questions and discussion
Papers…
luizdasilva.wordpress.com
On email…
dasilval@tcd.ie
ldasilva@vt.edu

28. Backup slides

29. Coordination for heterogeneous and multi-hop networks
• Distributed spectrum sharing for multihop topologies and HetNets (relays,
coexistence between small and large cells)
• Adaptations: channel selection, transmit
power
• Goals: network-wide spectrum efficiency,
fairness, network connectivity, coverage
• Cooperative game theory, coalition
formation
Types of coalition in equilibrium as a
function of link range
Z. Khan, S. Glisic, L. A. DaSilva, and J. Lehtomaki,
“Modeling the Dynamics of Coalition Formation Games
for Cooperative Spectrum Sharing in an Interference
Channel,” IEEE Trans. on Computational Intelligence and
AI in Games, 2011
J. E. Suris, L. A. DaSilva, Z. Han, A. B. MacKenzie, and R.
S. Komali, “Asymptotic Optimality for Distributed
Spectrum Sharing Using Bargaining Solutions,” IEEE
Trans. on Wireless Communications, Oct. 2009