Multiuser Receivers for Synchronous CDMA Systems

1
Multiuser Receivers for Synchronous CDMA Systems

• Kiran Puttegowda
• and
• Soshant Bali

2
Agenda

• Motivation
• Direct Sequence Spread Spectrum
• Multiple Access Interference
• Multiuser Detection Techniques
• Optimal
• Interference Cancellation
• Decorrelator
• Minimum Mean Square error
• Future Work

3
Motivation

• Cellular CDMA systems have many advantages over
  T/FDMA systems
• Resistance to multipath Fading
• Increase in System Capacity
• Eliminates need for frequency planning (Reuse
  Ratio 1)
• Soft Handoffs decreases call dropping
  probability
• Graceful degradation Soft limit on number of
  users
• May take advantage of low voice activity of
  normal speech
• Wideband CDMA proposed for Third Generation
  Wireless Personal Communications

4
Direct Sequence Spread Spectrum System
5
Direct Sequence Spread Spectrum
6
DSS System
7
DSS System
8
Multiple Access Interference
9
Multiple Access Interference
10
Multiple Access Interference
11
Multiple Access Interference
12
Multiple Access Interference
13
Near Far Effect
14
Multiple Access Interference Simulation
15
Multiuser Detection

• Demodulate the signals of all users at the same
  time
• Use knowledge of chip sequence, and power levels
  of other transmitters to jointly detect all
  users signals

16
Successive Interference Cancellation

• Subtract off the contributions of Multiple Access
  Interference from the received signal r(t)
• Cancel the strongest signal first
• It has the most negative effect (maximum benefit
  first)
• Best bit decision can be made on that signal
  (most reliable cancellation)

17
Successive Interference Cancellation
18
Parallel Interference Cancellation

• Generate estimates of the users signal using
  correlators
• Reconstruct estimates and subtract off from the
  received signal r(t)
• Decision statistic is then computed using this
  estimate

19
Parallel Interference Cancellation
20
Optimal Multiuser Receiver

• Maximum Likelihood Sequence Detector
• Similar to decoding of convolutional codes or
  GMSK demodulation
• Impractical complexity per bit decision is 2(k)
• Considered a great theoretical feat
• Sub-optimal solutions sought

21
Optimal
22
Sub-optimal techniques

• Linear
• Decorrelator and MMSE
• Non-linear
• Parallel and Successive Interference Cancellation
• Linear receivers
• computationally intensive and numerically
  sensitive
• Non-linear
• easy to implement and usually done in many stages

23
Decorrelator

• A linear transformation
• C correlation matrix of users pn-sequences
• Removes correlation between the elements of
  matched filter outputs
• Degrades as cross correlations increases
• Complex inverse matrix calculation
• Noise enhancement

24
Decorrelator
25
MMSE

• Minimizes mean square error
• Linear transformation
• No noise enhancement
• Performance depends on power of interfering users

26
MMSE
27
Comparison SNR
28
Comparison No. of Users
29
Future Work

• Simulation of Near far effect Some signals are
  received at a higher power level than others
• Validation of Simulation results with analytical
  values