Docstoc

Adaptive Signal Processing Research in Wireless Communications Anders

Document Sample
Adaptive Signal Processing Research in Wireless Communications Anders Powered By Docstoc
					       Research in
Wireless Communications

     Anders Høst-Madsen
       Department of EE
      University of Hawaii

                             1
                             Overview

   The wireless revolution(s)
   Wireless systems
   Wireless principles
   Error correcting codes
   Information theory
                                ALOHA net

   University Hawaii pioneer in wireless networking
   World’s first wireless “internet”
   Part of DARPA net project
        Military project developing internet
   Developed by Norm Abramson in 1970s
   Interisland wireless connections
   ALOHA protocol
        Still widely used
         The Three Wireless Revolutions

   The Wireless
        Marconi 1899
   Wireless Everywhere – The past
      Radio in every home
      1930s
   Wireless Everyone – The present
      Mobile phones
      1990s
   Wireless Everything – The future
      Devices
      2000s, 2010s
                  Mobile Communication
                       World Wide

   Europe
      Largest mobile phone market
      Driving innovation
      Rather low internet penetration
   North America
      High internet penetration
      Low mobile phone penetration
   Japan & Korea                              Even babies in Korea have mobile phones!
        High penetration
          – 50% in Japan
          – 70% in Korea!
      Mobile phones are small!
      First with IMT2000 (2000!)
      Japan behind with (fixed) internet
      Japan: 6 mill wireless internet subscribers (I-mode)
  Wireless Everywhere
Broadcast Radio Systems
Wireless Everyone
Cellular Systems
                     Wireless Everything

   Every device can have built in wireless transceiver
        TV, HiFi, speakers
          – Home entertainment center
        Watches
          – Microsoft watch
        Refrigerators
        Plants
        Pacemakers and other internal sensors
        Clothes: environment, sunshine
        Microwaves
   RFID technology
              Wireless Sensor Networks

   Sensors
      Temperature, humidity
      Movement
      Chemical, biological detectors
      Cameras, microphones
         – X10
   Actuators
      Acts on sensed data
      Robots
      Computers
   Applications
      Home environment, cleaning,
      Environment monitoring
      Battlefield, homeland security
Protecting Endangered Plants

                Hard to do without knowing plants'
                needs, so
                Study the environment
                See how the plants are doing
                Use sensors and hi-res cameras
Wireless Everything
 Ad-hoc networks
               Communication System


0110100011100110…
                                                0110100011100110…

                           Bit errors


0110100011100110…
                                                 0110000011101110…


     Errors at random positions
     Bit error rate (BER)=percentage of wrong bits
               Error Correcting Coding

Repetition coding



   01101   000 111 111 000 111

                                            100 101 111 110 111

Parity check                                0   1   1   1   1

   01101       011011      00101   011010

Hamming code
                      Information Theory

   n repetitions
      Error decreases to zero with increasing n
      Rate decreases to zero with increasing n
   Shannon
      It is possible to communicate with zero BER with a finite rate!
      Channel capacity

           R  C ( p)  1  p log 2 p  (1  p) log 2 (1  p)
           C (0.1)  0.53
           C (0.01)  0.92
           Research in Communications

   Coding
        Shannon(1948): proved theoretically that communication near channel
         capacity possible
        45 years of efforts to practically approach Shannon limit
        1993: turbo codes, close to Shannon limit
        1995: LDPC codes rediscovered, invented in 1962
        Currently: improvement of turbo, LDPC codes
   Information Theory
        Channel capacity of more advanced networks
                     Ad-Hoc Networks
                   Existing Technologies

   IEEE 802.11
        Not real ad-hoc standard
   Bluetooth
      Primitive
      For small networks
Wireless Everything
 Ad-hoc networks
Wireless Everything
 Ad-hoc networks
              Other Systems Research

   Signal Processing
   Control Theory
   Networking
   Image Processing
                                                        Volumetric Data Compression using the
                                                           Derivative of Gaussian Transform
                                                                                                                                                                    Professor Todd R. Reed




Original                                                                                       Lossless                         Compressed
                            3-D DGT            Quantization
volume                                                                                         Coding                           volume

                                                                                      1
           n= 0         n= 3          n = 10   n = 21                                                                             CSF
                                                                                     0.9                                          Basis function sensitivity

                                                                                     0.8

                                                                                     0.7




                                                              Contrast sensitivity
                                                                                     0.6

                                                                                     0.5
           n = 36       n = 53        n = 74   n = 99
                                                                                     0.4

                                                                                     0.3

                                                                                     0.2

                                                                                     0.1

                                                                                      0
                       A 1-D DGT basis.                                                    0     5   10   15     20     25     30     35     40      45        50
                                                                                                            Frequency in cycles/degree
                                                                                                                                                                       A 3-D basis function.




                                                                                                                                                                     Data courtesy of:




                    The original volume.                                                             After 30:1 compression.

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:5
posted:2/8/2012
language:
pages:20