50 users per cell (NU=600) NT=6

1
Multi-Cell User Scheduling and Random Beamforming
Strategies for Downlink Wireless Communications
Xiaojun Tang, Sean A. Ramprashad, and Haralabos
Papadopoulos
DOCOMO USA LABS
Background and Motivation
A-B-C Scheme Animated
Overview of Schemes
Results

• Simulation Layout
• 12 cells separated by 1 unit on a topology
  wrapped to form a flat 2D torus.
• All cells are equivalent on such a torus.

• Mitigating inter-cell interference (ICI) in
  multi-cell downlink is critical to increasing
  throughputs in wireless systems.
• It is a topic of much interest in research and in
  standards, e.g. in Coordinated Multi-Point (CoMP)
  in LTE/LTE-A.
• Strategies for SISO include frequency-reuse and
  power control.
• MIMO provides a new dimension of space.
• One can leverage space by joint MIMO processing
  across multiple base stations (BS), e.g. by
  Network-MIMO
• Can remove inter-cell interference within
  clusters of cells.
• However, it can require a high degree of
  coordination, e.g.
• Full channel state information (FCSI) feedback,
• Full data sharing/signaling within clusters of
  cells.

Schemes Interference Management Mechanism
Fully Independent Scheduling No interference management
TDM Scheme Avoid strong interference by dividing transmission resources
Multi-Cell SINR Feedback Mitigating interference via user selection with multiple beam options and pilot sensing across cells
A-B-C Scheme Mitigating interference via joint user selection and joint beam coordination across cells with limited direct information exchange

• 0) Assign cells to an A, B or C subset
• Cells in each subset act independently and are
  separated geographically.
• Shuffle 1

a1
a2
a3
a4

• System uses frequency reuse factor 1. This is
  not a frequency reuse pattern.

1) Baseline/Independent Schemes

• Fully Independent System
• Cell operate as isolated cells with frequency
  reuse 1.
• There is no exchange or use of inter-cell
  information.
• TDM/FDM System
• Like fully independent but with frequency reuse
  3.

• 1) A subset sends pilots and gets CQI from
  users
• B and C cell users sense pilots
• A users and beams are scheduled

• 2) B subset sends pilots and gets CQI from
  users
• Get FCSI or CQI from scheduled A users
• B users and beams are scheduled

• Per-User Throughput (Empirical CDF)

Goal

• NT 6, 50 users per cell, 1 beam/cell,
• The TDM scheme performs the worst due to loss of
  degrees of freedom.
• Also the scheduler does not know ICI.
• The Fully independent scheduling scheme
  increases per-user throughput despite no
  interference management.
• Also the scheduler does not know ICI.
• The Multi-cell SINR feedback scheme performs
  well for users relatively close to BSs.
• The ABC scheme performs the best, especially for
  edge users.
• Beam coordination benefits edge users greatly.

• We would like to consider strictly cellular
  transmission that
• Aligns interference across cells by exploiting
  the spatial dimension without joint-cell
  transmission.
• Requires minimal intra/inter-cell information
  exchange.

2) SINR Scheme
users sending feedback
SINR Joint Multi-Cell User Scheduling by
Sensing Pilots
illustration of beam pilot
1) Each BS selects m beams for its cell randomly
and independently, and sends pilots.
2) Each user estimates beam CQIs, selects a beam
from its BS, sends the selection and SINR to
its BS .
3) Each BS selects its user based on rate and
beam requests and transmits.
System Model
Fig 4.

• Consider a multi-cell downlink system consisting
  of
• NB base stations (BSs), each with NT transmit
  antennas.
• Each user is served by one (and only one) BS.

• Per-Cell Throughput

• 3) C subset sends pilots and gets CQI from
  users
• Get FCSI or CQI from sched. AB users
• C users and beams are scheduled

• 4) All cells transmit at same time with frequency
  reuse 1
• A cell users tend to get higher rates than B and
  C cell users, and so on...

• Fig. 1 The received signal of user k (in cell
  0) at time t is
• hk,i is FCSI from BS i to user k
• xi is the transmission from BS i.
• The FCSI accounts for path loss, shadowing and
  block Rayleigh fading.

illustration of data transmission
illustration of beam pilot
illustration of data transmission
users sending feedback
Fig 6.
Fig 5.
50 users per cell (NU600)

NT6

3) A-B-C Multi-Step Scheme
Observations

• 5) Shuffle A, B or C subset
• Then go to Step 1 with new assignment.
• Shuffle 2
• Shuffle 3

• Cell Partitioning
• Cells are partitioned into A, B and C
  subsets
• Frequency reuse 1 is used.
• Multi-stage scheduling
• Limited Inter-Cell Beam Coordination
• Cells operate with CQI feedback with following
  additions
• B cells get FCSI or CQI from scheduled A cell
  users.
• B cells select beams to limit ICI to A cells
• C cells get FCSI or CQI from scheduled A
  B users.
• C cells select beams to limit ICI to B and
  C cells
• A/B/C assignments are shuffled to preserve
  fairness.

• Increasing the number of beams per cell
  increases intra-cell interference,
• Can hurt schemes without any interference-avoidanc
  e mechanism.
• Increasing NU benefits the SINR and ABC schemes
  via user-selection
• Typical of opportunistic beamforming, but SINR
  has limits vs NT.
• The ABC scheme, with beam coordination, does
  leverage increased NT .

• Linear Opportunistic (Random) Beamforming
• mi data streams sj,i j 1,,mi are multiplexed
  at BS i
• where qj,i is the (random) beam vector
  assigned to stream sj,i .
• sj,i is given power 1/mi,
• The Signal to Noise and Interference Ratio (SINR)
  for user k using beam j from BS 0
• Achievable rate
• Opportunistic beamforming requires limited
  feedback
• No FCSI is needed, only channel quality
  information (CQI) in the scalar forms of
  .
• Multi-user Proportional Fair Scheduling (MPFS)

a1
a2
a3
a4
Summary

• Considered limited coordination (CQI only or
  limited FCSI) random beamforming strategies for
  multi-cell downlink.
• Opportunistic beamforming does have limits in
  leveraging the spatial dimension for managing
  inter-cell interference
• Changing NT, without increasing pool of beams and
  using joint beam selection, has little affect on
  ICIs (Fig. 5)
• To leverage the spatial dimension, some joint
  inter-cell beam selection is required as in A-B-C
  scheme
• We considered a joint beam scheme which requires
  some limited exchange A cells?B,C cells A,B
  cells ?C cells.
• CQI could be used in place of Full CSI in the
  A-B-C scheme.
• Requires larger beam pools and possibly more
  pilots.

c1
c2
b1
b2
c3
c4
b3
b4