The First NRTA Knowledge Dissemination and...

4G++: Advanced Performance Boosting Techniques in 4th Generation

Wireless Systems and Beyond

Khaled Fouad ElsayedCairo University

The First NRTA Knowledge Dissemination and Networking

ConferenceOctober 12, 2011

1

Khaled Fouad Elsayed

Agenda

� Mobile Communications Outlook

� LTE and LTE-Advanced

� The 4G++ Project Scope

� Research Areas and Outputs

� Potential for Applications in Egyptian Market

� Discussion

October 11, 2011 2


Mobile Communications Outlook

� The Mobile Tsunami is Near� In 2015, the average mobile user will consume 1,118 megabytes of traffic per month (compared with 65 megabytes today)

� Exabyte = 1 Gig Gigabyte, 10^18, 2^60

October 11, 2011 3


Video is the Killer Application

� Mobile video is forecast to represent 66 percent of all mobile data traffic by 2015, increasing 35-fold from 2010.

� Mobile traffic originating from tablet devices is expected to grow 205-fold from 2010 to 2015.

� Global mobile data traffic increased 159 percent from calendar year 2009 to calendar year 2010 to 237 petabytes per month.

October 11, 2011 4

� Global mobile data traffic grew 4.2 times as fast as global fixed broadband data traffic in 2010.


Mobile Data Traffic Surpasses Voice

� Mobile data surpassed voice on a global basis during December of 2009.

� Some 400 million mobile broadband subscriptions generating more data traffic than the voice traffic from the total 4.6 billion mobile subscriptions.

� Ericsson consumer insights study showed that as much as 80% of mobile broadband users demand anytime, anywhere access.

� As mobile data traffic increases, the mobile broadband connection is becoming as personal as the mobile phone..

� Mobile networks will need to be dimensioned for data as well as voice � significant network modernization and

transformation to all-IP technologies and new support systems.

� � Enter LTE/LTE-A.

October 11, 2011 5


Opportunity Knocks Again!

� We are in the midst of the fifth major technology cycle of the past half a century.

� The era of the Mobile Internet

� Mary Meeker of Morgan Stanley“Rapid Ramp of Mobile Internet Usage Will be a Boon to Consumers and Some Companies Will Likely Win Big (Potentially Very Big) While Many Will Wonder What Just Happened.”

October 11, 2011 6


Wireless Technology Must Scale

� Or else the cycle is broken ☺

� ITU Definition of 4G: Very stringent requirements on rate and mobility.

� Neither IEEE 802.16e nor LTE Rel. 8 satisfies IMT-Advanced requirements.

� LTE-Advanced and IEEE 802.16m efforts were initiated to meet the targets � True 4G Wireless

October 11, 2011 7


LTE Requirements

� Significantly increased peak data rates

� Improved spectrum efficiency

� Improved latency

� Scalable bandwidth 1.4 up to 20 MHz

� Acceptable system and terminal complexity, cost and power consumption

� Compatibility with earlier releases and with other systems

� Optimised for low mobile speed but supporting high mobile speed

October 11, 2011 8


LTE Main Technologies

� OFDMA/SC-FDMA

� MIMO (Spatial Multiplexing, Diversity, Beam Forming)

� Adaptive Modulation and Coding

� Turbo-codes

� Hybrid ARQ

� Channel-aware Radio Resource Management

� Quality of service support

� [Release 8]

October 11, 2011 9


LTE-Advanced

� All of LTE plus� 1 Gbps downlink and 200 Mbps uplink (wireless!!)� 100 MHz bandwidth

� 8x4 MIMO

� Carrier-aggregation in case no continuous 100 MHz band available.

� Relays

� Coordinated multipoint

� [Release 10 and above]

October 11, 2011 10

Khaled Fouad Elsayed 11October 11, 2011

Multiple Antenna Transmission

� SISO

� SIMO

� MISO

� MIMO� Diversity Mode: space-frequency block codes �Good for users with bad channels to combat low SNR

� Beam forming and precoding: focus power in certain direction (SDMA).

� Spatial multiplexing: each antenna transmits a different stream � Good for good channels

� LTE-A: 8x4 MIMO is supported


The 4G++ Research Project

� An ambitious project tackling a multitude of radio resource management techniques with goal of enhancing spectral efficiency in terms of bits/sec/Hz in LTE/LTE-A.

� Classical single-cell and multi-cell (resource allocation/power control).

� More challenging scenarios:� Carrier-aggregation� CoMP� D2D Underlay� Femto/PicoCells� Relays� HetNet

� Self-Organizing Networks (SoN) Concepts.

October 11, 2011 12


4G++ WP Structure

13

WP3: Channel-aware RRM and Scheduling

WP4: Inter-Cell Interference Coordination

WP2: Development of Accurate

LTE/LTE-Advanced System Model

WP6: Dissemination

Dissemination, Exploitation, Standardization

WP1: Project Management

Planning, monitoring, reporting

WP5: Relaying and Network Coding in LTE-Advanced

October 11, 2011


Research Area 1: Channel-Aware RRM and Scheduling

� Objectives:

1) To investigate, propose, and evaluate schemes for channel-aware scheduling in uplink directions in LTE/LTE-Advanced

2) To design effective schemes for radio resource management (RRM) for Coordinated Multipoint (CoMP) transmission.

3) RRM for Carrier-Aggregation Setups

14October 11, 2011


OFDMA

� Multiple access scheme based on OFDM and enjoys inherent characteristics of immunity to inter-symbol interference and frequency-selective fading.

Symbols (Time)

Symbols (Time)


Resource Blocks


Riding the Peaks!

� Allocate the common resources to the user that can bestexploit it � Maximize

throughput at the cost of unfairness

� Opportunistic Scheduling:Exploits the time-varying channel conditions and maximizes the system performance stochastically under a certain fairnessconstraint.

� Opportunistic scheduling with fairness in users’

packet delays is a very challenging issue


LTE Uplink RRM

� Integrated Framework for LTE resource blocks allocation (number and location), power adaptation, and MCS selection.

� QoS-aware

� Interference-friendly (multi-cell taken into account but not explicitly).

18October 11, 2011


Coordinated Multi-point (CoMP)

� A CoMP refers to a system where several geographically distributed antenna modules coordinate to improve performance of the served users in the coordination area

� RRE/RRH: Remote radio equipment/head connected with Radio over Fiber (RoF) to base station (eNB)

� Multi-cell coordinated transmission:

� Coordinated Scheduling (one at a time).

� Joint Processing/BF

� Multi-cell reception considered for beyond Rel. 10


CoMP RRM

� Current state of the art is that the serving eNB selects resource blocks and power and provides this information to the cluster of eNB’s that are collaboratively transmitting to the UE.

� Allocation decision based on max sum rate seen collectively by the cluster not just the serving eNB.

22October 11, 2011


CoMP RRM

� Formulation based on SLNR

� Both Coordinated Scheduling and Joint Processing Approaches

� Different Criterions for reaching max-sum-rate.

23

20 25 30 35 4020

30

40

50

60

70

80

Number of users per cell

Throughput per RB [Kbps]

CS based on SLNR and leakage Threshold

CS based on SLNR and look ahead

CS based on SLNR and marginal utility

CS based on SINR

October 11, 2011


Carrier Aggregation

� To meet the IMT-ADV requirements for 100 MHz system bandwidth, LTE Release 10 introduced CA as one of the main features to scale bandwidth beyond 20 MHz.

� Two or more component carriers of the same or different bandwidths are aggregated to support wider bandwidths in the DL or UL. 5 CC � 100 Mhz, 5 PHY’s

(depends on UE’s capability, significant challenge)


Research Area 2: ICIC

� Objective of work package 4 is to devise efficient schemes for RRM in a multi-cells scenario. ICIC = Inter-cell Interference Coordination 1) Efficient solution of the multi-cell radio resource

management as an optimization problem

2) The design of autonomous allocation scheme based on limiting interference contribution of each base-station

3) Using interference alignment technique as a means for increasing systems capacity

4) Autonomous RRM (and eICIC) for femto-cells and in general in Heterogeneous Networks (HetNet).

25October 11, 2011


Interference Alignment (IA)

� Interference alignment scheme is found to achieve K/2 DoF for K interfering TX/RX pairs [Cadambe and Jafar, ‘07]

� Interference as a friend not foe!

26


IA in D2D Uderlay Cellular Networks

27

Control Links

� UE units may communicate directly with each other over the D2D links.

� The UE in D2D connections remains controlled by the eNBsand continue cellular operation.

� Reuse of resources between cellular and D2D and among D2D clusters/groups.

5G Technology ☺☺☺☺

October 11, 2011


IA in D2D underlay Cellular Networks

28

� Grouping of D2D Pairs� Position-based

� Channel-based

� Distance-based

� Resource block allocation to groups.

Channel Based Grouping

October 11, 2011


IA in D2D underlay Cellular Networks

29

� Excellent gains compared to P2P.

� Gain increases with SINR increase.

� Staggered-IA or less CSI information

October 11, 2011


Channel Estimation Error Effects on IA Performance

30

MIMO M × 2 X Channel

October 11, 2011


More Work on IA

� Staggered/Blind IA

� IA with limited cooperation

� Scheduling and resource allocation with IA/OFDMA.

31October 11, 2011


Other Work on ICIC

� Classification of ICIC (static and dynamic)

� ICIC as optimization

� Autonomous ICIC and power control

� eICIC (femto+pico)

� Femto Operation in SoN manner

� Femto PCI Confusion Problem

� IA theoretical work

32October 11, 2011


Research Area 3: Relays

� Potentials of Relaying

� Enhance Throughput.

� Extend Coverage.

33October 11, 2011

Khaled Fouad Elsayed October 11, 2011

Relays in LTE-A

34

� Type-1 RN: It is an in-band RN

� It controls cells, appearing as a new cell to the UEs. Each of its cells have their own Physical ID, SYNCH channels, reference symbols, etc.

� It appears as a Release 8 eNB to Release 8 UEs, for backward compatibility.

� Type-1a has the same characteristics of Type-1, but operates out-of-band.

� Type-2 RN

� It does not have a separate cell ID and thus does not create any new cells.

� It is transparent to LTE UEs

� At least "Type-1" and “Type-1a” RNs are supported by LTE-Advanced


Proposed Research in LTE-A Relays


Relay Performance

36October 11, 2011

Without Relays With 2 Relays per Sector


Relay Performance

37October 11, 2011

Where to Place Relay (R, Θ)


Heterogeneous Networks (HetNet)

� Radio link performance is approaching theoretical limits with 3G enhancements and LTE.

� Next performance leap in wireless networks will come from the network topology. � Shrinking reuse distance of wireless resources. � Improving bits/sec/Hz/ per unit area.

38October 11, 2011


HetNet: Challenges with Femto Deployment

• Unplanned deployments of Closed Access femtoscreates significant interference scenarios:

� DL of macro UE can be jammed due to closeproximity to femto.

� UL transmissions from macro UE can severelyimpede femto.

Femtos (HeNBs) with Restricted Association


Project Team

� PI: Khaled Fouad Elsayed, Cairo Univ.

� Lead Researchers:

� Mostafa Elkhmay (Alex Univ/E-Just)

� Mahmoud Hamed Ismail (Cairo Univ)

� Haitham Hamza (Cairo Univ)

� Research Assistants:

� 8+ (growing)

� 5 (pursuing masters, working elsewhere)

40October 11, 2011


Opportunities for Collaboration (1)

� The Trio Mobile Operators should be aware of inherent complexities in 4G Network Planning.

� Reuse-N is obstacle to capacity.

� Full dynamic solution may not offer predictable performance.

� Cell-edge, cell-edge, cell-edge.

� Energy-efficiency is major issue.

� HetNet: even more complex.

� SoN is a key enabler.

41October 11, 2011


Opportunities for Collaboration (2)

� Developers of Protocol Stacks and System Developers

� All RRM functions are left open in the standard.

� Key product differentiator.

� SoN capable stacks with proven capabilities are highly required in the market.

42October 11, 2011


Conclusions

� Good time to jump on the bandwagon of an exciting technology in the midst of the fifth major technology cycle.

� Egypt’s youth is capable of doing good and timely research.

� Team building is most challenging (getting good people who believe in research).

43October 11, 2011


Questions?

[email protected]

www.4gpp-project.net

44October 11, 2011

The First NRTA Knowledge Dissemination and...

Documents

Transcript of The First NRTA Knowledge Dissemination and...