MBMS Evolution in Release 7 and Beyond

1
MBMS Evolution in Release 7 and Beyond Qualcomm
Germany
2
MBMS evolution in the scope of C-MOBILE

• IST Project C-MOBILE (MBMS for the Future Mobile
  World) investigates enhancements to MBMS for 3G
  and B3G systems
• Areas of interest RAN, CN, Content, Business
  Models

WP2Requirements and Business Models

• Scope
• RAN enhancements for MBMS with reference to 3GPP
  HSPA LTE, and Beyond-LTE systems
• Candidate concepts
• MBSFN operation w/ advanced receivers
• Hierarchical modulation
• Relaying strategies for MBMS
• Optimized scheduling and switching

WP3RAN Enhancements
WP4CN Enhancements
WP5Applications and Content
WP6Proof of Concepts
3
WCDMA HSPA evolution in 3GPP Release 7

• Higher-order modulation (DL 64-QAM, UL 16-QAM)
• MIMO (peak rate 28Mbps Rel7, 42Mbps Rel8)
• CPC Layer-2 enhancements
• Multicarrier operation (N-fold data rate
  increase)
• Multicast/Broadcast Single-Frequency Network
  (MBSFN) operation
• Tries to overcome the cell-edge problems of MBMS
• Goal significant reduction of inter-cell
  interference
• Reuse of existing channel structures MTCH / FACH
  / S-CCPCH
• 16-QAM modulation for S-CCPCH
• Advanced receivers
• UE receive diversity
• Linear equalizers (G-RAKE, LMMSE)

4
MBSFN for WCDMA

• What is MBSFN?
• Use of a common scrambling code and the same
  spreading codes to simultaneously broadcast the
  same MBMS channels in a number of cells that form
  an MBMS cluster
• It promises to significantly improve spectral
  efficiency
• Interest / Challenges
• Conditions under which MBSFN leads to
  improvements in spectral efficiency, compared to
  Rel6 MBMS with MRC of several RLs?
• Implication of MBSFN on mobile handset receivers?

5
MBSFN design considerations

• Design considerations for MBSFN PHY (TR 25.905)
• Configuration of a common scrambling code
• Base site synchronization
• Receiver support for sufficient delay spread (up
  to 33us)
• Suitable UE equalizer
• UE receive diversity
• Higher-order modulation on S-CCPCH (16-QAM)
• Flexibility of resource partitioning between
  unicast and MBSFN
• Resource partitioning two options
• Time-multiplexing of MBSFN and unicast traffic
  (no need for dedicated carrier)
• MBSFN on an auxiliary carrier, both applicable to
  TDD and FDD(dual-receiver terminal has
  advantages w.r.t. of signaling overhead)

6
Channel Interference characteristics

• Conventional MBMSBS specific CIRs to the UE
  (MRC)

• MBSFNJoint CIR between BS sites and UE

BS 1
time
Propagation delay
BS 2
BS 123
time
time
Propagation delay
Propagation delay
BS 3

• Identical signals are transmitted (more or less)
  synchronized from all BS sites within the MBMS
  cluster

time
Propagation delay
7
Impact on UE receivers

• CIR structure
• Long effective CIRs
• Even longer CIRs when the BS sites are only
  coarsely synchronized
• Interference structure
• High multi-path interference spread over a large
  window
• Virtually no inter-cell interference inside the
  MBMS cluster
• Border effects inter-cell-interference from rest
  of the world
• MBSFN promises very high receiver output SNR
• System may become code-limited rather than
  interference-limited!
• Transport format optimization allows to maximize
  spectral efficiency
• How should the equalizers be timed and adapted?

8
System-Level Simulation Methodology

• Comparison of conventional MBMS (Rel6) vs. MBSFN
  under the same prerequisites
• Scenario of 19 sites with 3 sectors each on a
  hexagonal grid, spacing 1200 meters
• 3GPP system-level spatial channel model (SCM)
  Urban Macro
• Receive diversity (2 UE antennas)
• UE LMMSE equalizers with 20 chip eqx window at
  2-times over-sampling
• SNR maximizing equalizer timing
• SNR loss from equalizer adaptation based on a
  training of 6 WCDMA slots (4 ms)
• Power allocation 90 for MBMS, 10 for the pilot
  
• Different S-CCPCH transport formats all with 64
  kbps payload data rate
• Performance criterionThe number of 64 kbps MBMS
  channels can be supported with a coverage of 95
  at a block error rate (BLER) not higher than 1.

9
Transport Format Options

• S-CCPCH transport formats (TF) of length 80 ms
  (120 WCDMA slots)with 64 kbps payload data rate
  (AWGN reference)

Ec/N0 for BLER 1
10
TF Optimization MBMS channels in 5 MHz

• The different Ec/N0 requirements of the transport
  formats allow to maximize the total of S-CCPCHs
  at 64 kbps for a given chip-SNR

• Chip-SNR Sum TFk(Ec/N0)k
• S-CCPCHs Sum TFk

11
Results for different RX schemes (at 5 outage)

• Chip-SNR vs. supportable S-CCPCHs

• CDF of Chip-SNR at equalizer output

12
Impact of coarse BS sync on different Eqx schemes

• Perfect BS synchronization

• Gaussian async. of 12 chips std.dev.

13
Conclusions

• MBMS Rel7/8 offers high spectral efficiency gains
  via MBSFN
• Equalizer design and adaptation of importance
• Accurate BS synchronization needed, coarse sync
  may reduce the benefit of MBSFN
• UE receive and macro-diversity extremely
  beneficial (outage design)
• Spectral efficiency gains in the order of factor
  2-5 compared to single-link reception and MRC
• Other C-MOBILE concepts may offer alternative
  opportunities for improvement, such as
• Hierarchical modulation (cf. DVB-T, MediaFLO)
• Relaying for MBMS

14
Thank you