Transmitter Design in Communication Systems - PowerPoint PPT Presentation

1 / 17
About This Presentation
Title:

Transmitter Design in Communication Systems

Description:

Encoding with CSI in Transmitter. Possible in slowly fading channels. ... Transmitter (Needs CSI). Correlation decreases the diversity gain. ... – PowerPoint PPT presentation

Number of Views:212
Avg rating:3.0/5.0
Slides: 18
Provided by: ece11
Category:

less

Transcript and Presenter's Notes

Title: Transmitter Design in Communication Systems


1
Transmitter Design in Communication Systems
  • Majid Fozunbal
  • October 19, 2001

2
Outline
  • Mathematical insight to communication systems
  • General scheme
  • Channel
  • Transmitter
  • Transmission over AWGN channel
  • Traditional method
  • Coded modulation
  • Non-ideal Gaussian channel
  • Transmission over fading channels
  • Encoding without CSI in transmitter
  • Encoding with CSI in transmitter
  • Diversity
  • Multiple antenna system

3
General Scheme
4
Channel
  • Distortion a linear operator.
  • Example any causal linear filter.
  • Having null space, not invertible.
  • Injective, one to one, preserves dimensionality.
  • Finite dimension, matrix theory.
  • Perturbation
  • Noise, interference.
  • Uncertainty.
  • No overlap, error-free inverse mapping.
  • Amount of overlap, error rate.

5
Channel (Cont.)
Channel Output Space
RN
  • RSN and RN are normalized per dimension (infinite
    dimension).
  • Gaussian noise.

6
Transmitter
  • An injective mapping from message space to signal
    space.
  • Compensating the channel distortion.
  • Finite dimension, soft balls, non-zero error
    rate.
  • Tradeoff between spectral efficiency and error
    rate defined by region of decision.
  • Infinite dimension, rigid spheres, zero error
    rate.
  • Increasing dimension, hardening the balls,
    increasing the spectral efficiency, decreasing
    robustness to noise variance.

7
Transmission over AWGN Channel
  • No distortion, just additive white Gaussian
    noise.
  • Band limited channel.
  • Constraint Power.
  • Goals
  • Given an error rate, maximize the spectral
    efficiency.
  • Given a spectral efficiency, minimize the error
    rate.

8
Traditional Method
  • Separate coding and modulation.
  • Adds redundancy to maximize the minimum free
    Hamming distance.
  • Information rate K/N.
  • Decreases the spectral efficiency.

4K sequences (4-PSK)
4K out of 4N sequences
9
Coded Modulation
  • Combined coding and modulation.
  • TCM, Multi-level coding, Coset codes.
  • Adding redundancy by using a bigger signal
    constellation.
  • Convolutional coder and nonlinear mapping.
  • Maximizing the minimum free Euclidean distance.

4K Sequences (4-PSK)
4K out of 8K (8-PSK)
10
Non-Ideal Gaussian Channel
  • Non-ideal AWGN channels suffer ISI.
  • Dividing to a bunch of ideal sub-channels.
  • Multi-carrier, OFDM.
  • Spectral water-filling maximizes power
    efficiency.
  • Bit-loading to optimize spectral efficiency.

Power allocation for sub-channels
11
Transmission over Fading Channel
  • Time variable operators.
  • Characterized by PDF (Rayleigh, Rician).
  • Coherence time, coherence bandwidth.
  • Flat fading, slowly fading.
  • Channel estimation.
  • Fading decreases the channel capacity.

12
Encoding without CSI in Transmitter
  • Ergodic channel.
  • Time interleaving to achieve capacity.
  • Infinite length interleaving is required.
  • Suitable for data.
  • Delay constraint, no good for speech.
  • Uniform power distribution.
  • Design criteria
  • Maximizing Hamming distance.
  • Product distance.
  • Note Performance depends on PDF.

13
Encoding with CSI in Transmitter
  • Possible in slowly fading channels.
  • Transmit more power and data when channel is
    good.
  • Temporal water-filling.
  • Bit-loading and variable size modulation skims.
  • Adaptive modulation TCM and Multi-level coding.

14
Diversity Techniques
  • Diversity changes the statistics of the received
    signal.
  • Time diversity.
  • Frequency diversity.
  • Space diversity
  • Receiver.
  • Transmitter (Needs CSI).
  • Correlation decreases the diversity gain.

15
Multiple Antenna Systems
X
H
Y
Y H X V
16
MIMO (NO CSI in Transmitter)
  • N by N MIMO.
  • N times of SISO capacity (no correlation).
  • Degrees of freedom
  • Combat fading.
  • Space-time TCM
  • High efficiency.
  • Complex decoder.
  • Space-time Block Coding
  • Maximum diversity.
  • Simple decoder.
  • Beam-forming.
  • BLAST Source separation.

17
MIMO (with CSI in Transmitter)
School of ECE
School of ECE
  • Beam-forming in transmitter and receiver.
  • Singular value decomposition (SVD).
  • A parallel bank of SISO channels.
  • Capacity is defined by PDF of singular values.
  • Temporal-spatial water-filling.
Write a Comment
User Comments (0)
About PowerShow.com