Wireless Communication Engineering Fall 2004

1
Wireless Communication Engineering(Fall 2004)

• Lecture 2
• Professor Mingbo Xiao
• Sept. 30, 2004

2
Announcements

• Homepage of the course http//eed.xmu.edu.cn/pers
  on/mbxiao/wireless_course/index.html
• Next class on Oct. 10 (Sunday)
• Homework 12 are due on Oct. 10
• Reading Chap. 2, 6, and 8 in textbook

3
Digital Communications

• Review the basic principles and concepts of
  digital communication systems
• Physical layer of the network protocol stacks
  (remember the OSI model?)
• Physical, Data Link, Network, Transport, Session,
  Presentation, Application

4
Elements of Comm. System

• Information Source and Input Transducer
• Source Encoder
• Channel Encoder
• Digital Modulator
• Channel
• Digital Demodulator
• Channel Decoder
• Source Decoder
• Output Transducer

5
Digital Communication System
6
Data Communication Terms

• Data - entities that convey meaning, or
  information
• Signals - electric or electromagnetic
  representations of data
• Transmission - communication of data by the
  propagation and processing of signals

7
Analog Signals

• A continuously varying electromagnetic wave that
  may be propagated over a variety of media,
  depending on frequency
• Examples of media
• Copper wire media (twisted pair and coaxial
  cable)
• Fiber optic cable
• Atmosphere or space propagation
• Analog signals can propagate analog and digital
  data

8
Digital Signals

• A sequence of pulses that may be transmitted over
  a medium
• Generally cheaper than analog signaling
• Less susceptible to noise interference
• Suffer more from attenuation
• Digital signals can propagate analog and digital
  data

9
Input Source

• Source output may be analog or digital
• Analog signal - signal intensity varies in a
  smooth fashion over time
• No breaks or discontinuities in the signal
• E.g., Output of Video or Audio
• Digital signal - signal intensity maintains a
  constant level for some period of time and then
  changes to another constant level
• E.g., Output of teletype machine

10
EM Spectrum for Telecom
Analog Signaling

• Figure 2.10 (pdf)
• Most spectra licensed 3G license is very
  expensive FCC is a mighty sector
• Infrared, ISM band, and amateur radio band are
  license-free
• HW1 Find out what spectrum is used for GSM,
  IS-95, 802.11b WLAN. What data rates are
  available in each system? What transmission
  characteristics makes these spectrum bands
  suitable for wireless communications?

11
(No Transcript)
12
Reasons for Choosing Data and Signal Combinations

• Digital data, digital signal
• Equipment for encoding is less expensive than
  digital-to-analog equipment
• Analog data, digital signal
• Conversion permits use of modern digital
  transmission and switching equipment
• Digital data, analog signal
• Some transmission media will only propagate
  analog signals
• Examples include optical fiber and satellite
• Analog data, analog signal
• Analog data easily converted to analog signal

13
Representation of Signals

• Time domain waveform
• amplitude vs time plot
• zero crossings relate to frequency content
• Frequency domain waveform
• amplitude vs frequency plot
• Related by Fourier Transform

14
Time domain
amplitude (volts)
time (seconds)
examining zero crossings suggests more than one
frequency present, with different amplitudes
15
Frequency domain
amplitude (volts)
f1
f2
frequency (hertz)
16
The Square Wave

• Waveform with equal on and off times
• Can be represented by a series of harmonically
  related sinewaves
• fundamental plus
• 1/3 third harmonic plus
• 1/5 fifth harmonic plus
• 1/7 seventh harmonic plus
• 1/9 ninth harmonic plus
• etc.

17
Square wave example
y sin(x) 1/3 sin(3x) 1/5sin(5x)
1/7sin(7x) 1/9sin(9x) 1/11sin(11x)
1/13sin(13x)
18
Square wave - power spectrum
implies infinite bandwidth to transmit a square
wave
19
Spectrum of a Data Pulse1
keep pulse width constant and increase
periodicity
20
Spectrum of a Data Pulse2
keep periodicity constant and decrease pulse
width
21
Coding

• Waveform Coding converts an analogue signal into
  digital form.
• Source coding modifies the analogue or digital
  source to make it best suited for transmission
  (e.g.MPEG / JPEG)
• Channel coding adds redundancy to improve
  detection / correction of errors
• Modulation coding modifies a modulation symbol
  set again to overcome errors

22
Analogue / Digital Conversion

• More grandly called pulse code modulation (PCM)
  involves
• Regular sampling of the input signal
• Conversion of the sample to a number
• Accuracy of the A/D converter depends on number
  of bits
• Speech 8 bit
• Music HiFi 16/18, even 24 bit
• Video 8 bit x 3 colours to give 24 bit

23
Nyquist Sampling

• Key goal - minimise number of samples for
  accurate representation of the signal?
• Nyquist criterion sample at twice the bandwidth
  of the signal (twice maximum frequency in a
  baseband signal)
• Sampling at less than this results in aliasing

24
Nyquist Sampling System
25
Nyquist Criterion

• Can be derived intuitively from our knowledge of
  Fourier series. The sampling process can be
  viewed as the mixing of the input signal with a
  train of very narrow data sampling pulses this
  results in a series of sum and difference
  components appearing at the mixer output for each
  harmonic of the pulse waveform.
• To reconstruct the input waveform we simply
  filter the output of the D/A converter.

26
Alias Effect
27
Aliasing

• If the sampling criterion is not met and we
  sample at less than the twice the maximum
  frequency of the input waveform, the sum and
  difference components associated with each
  harmonic of the input waveform overlap with those
  of adjacent harmonics and we can no longer
  separate out the sampled waveform by filtering.

28
Practical Sampling

• Need to filter out baseband components above the
  range of interest
• Use a sampling frequency of 2.2 fmax to allow for
  practical filters
• e.g. speech telephony 300Hz 3.4kHz is sampled
  at 8 kHz
• CD sampling at 44.1 kHz for 20 kHz audio band

29
Dynamic Range1
30
Dynamic Range2

• Important that A/D converter can deal with both
  large and small signals.
• Ratio of Vmax to Vmin over which converter will
  operate is its Dynamic Range
• Depends on the number of bits the converter uses
• More bits means more Quantization levels

31
Quantization Noise
32
Companding

• A technique for reducing the number of bits while
  achieving an equivalent dynamic range or signal
  to Quantization noise level
• COMPressing and expANDING
• Decrease Quantization step size for small signals
  and increase for large signals
• International standards for telephony
• A Law European
• m Law USA

33
Homework

• Find out the details of A Law and m Law
• Show that the peak S/N ratio for A/D conversion
  is given by 3M 2/2, where M2 n and n is the
  number bits.

34
Source Coding

• Source coding finds a digital representation of
  source messages with little or no redundancy.
• Also known as data compression.
• Usually, now, an algorithm to realise bit or
  symbol content compression in addition to
  (waveform) A/D conversion.
• Two types lossy and lossless compression
• Voice coding can be lossy, but some data file
  compression must be lossless.

35
Entropy and Optimal Coding

• Entropy of a discrete random variable X is
  defined as
• Entropy is a measure of uncertainty, also a
  measure of the information.
• Entropy is the lower bound of the average code
  length of the decodable binary coding.
• Heuristics shorter codeword for symbol occurs
  with higher frequency.
• The codes are of variable length.

36
Current Algorithms

• Image compression algorithms
• MPEG moving image
• JPEG static image
• Voice Music
• Less well standardised
• GSM codecs / MP3
• Complementary coder / decoder

37
Classes of Speech Coding

• Waveform coding i.e. A/D conversion
• Frequency domain coding vocoder
• Parametric coding
• e.g. LPC linear predictive coding
• Maintain the subjective quality of the signal
  (audio or visual) not, necessarily, the shape of
  the input waveform.

38
Voice Coding

• Use of companding in a speech band voice coder (8
  bit A or m law) gives the equivalent subjective
  quality as a 12 bit linear coder.
• Sampling at 8000 times per second gives the
  familiar 64 kbit/s rate
• At the exchange a number of these 8 bit words
  from different phone source are assembled
  (multiplexed) into a FRAME

39
Assembling a Frame
40
E Series Frame

• E European standard
• E series frame is 8 x 32 256 bits wide
• and is sent at 8000 times per second
• Bit rate of basic E1 frame is 256 x 8000 2.048
  Mbit/s
• HW3 What source coding techniques are used in
  GSM and IS-95?

41
Channel Coding

• Is applied to communication links to improve the
  reliability of the information being transferred.
• By adding additional bits to the data stream
  which increases the amount of data to be sent
  it is possible to detect and even correct errors
  at the receiver.
• Also known as Error Control Coding.

42
Basic Approaches

• Error detection involves recognising that part
  of the received information is in error and
  requesting a repeat transmission.
• CRC Cyclic Redundancy Check
• ARQ Automatic Repeat Request
• Error detection and correction possible with
  added complexity and without re-transmission.
• FEC Forward Error Correction

43
Parity

• How does ARQ detect errors?
• One of the simplest yet most frequently used
  technique is the parity check bit.
• e.g. modem set up includes odd and even parity
  alongside stop bits
• Parity Check is a single bit (1 or 0) added to
  the end of the data word such that the number of
  1s in the new data word is even for even parity
  or odd for odd parity.

44
Types of FEC Coding

• Two main types
• Block coding where a group (or block) of bits
  is processed as a whole in order to create a new
  (longer) coded block for transmission. There is a
  complementary block decoder in the receiver.
• Convolutional coding which operates on the
  incoming serial bit stream generating a real-time
  encoded serial output stream.

45
Basics of Block Coding

• The terminology for block coding is that an input
  block of k bits give rise to an output block of n
  bits this is called an (n,k) code.
• The increase in block length means that the
  useful data rate (information transfer rate) is
  reduced by a factor k/n this is known as the
  rate of the code.
• The factor 1 - k/n is known as the redundancy of
  the code

46
Hamming Code Example
47
Hamming Codes

• Named after their discoverer R.W. Hamming
• A rate 4/7 Hamming code as example
• Each of the 16 possible four-bit input blocks is
  coded into a 7-bit output block
• The 16 output blocks are chosen from 27 128
  possible seven bit patterns as being most
  dissimilar. Each differs by 3 bits.
• If 1 error occurs receiver can correct it
• If 2 errors occur receiver can recognise it
• If 3 errors occur they go undetected

48
Block Code Families

• Hamming codes are a subset of the more general
  code family known as BCH (Bose-Chaudhuri-Hocquenhe
  m) codes discovered in 1959 and 1960.
• Whereas Hamming can detect up to 2 or correct 1
  error general BCH codes can correct any number of
  errors if the code is long enough, e.g. (11,1023)
  can correct 255 errors used in deep space
  probes.

49
Interleaving

• Block codes work well where errors are
  distributed evenly fixed networks.
• Mobile radio errors occur in bursts as signal
  fades so we use interleaving.
• Implementation
• Read encoded data blocks in rows of matrix
• When matrix full, read out blocks as columns
• At receiver inverse process for de-interleaving
  which redistributes the burst errors uniformly
• Penalty latency / time delay for processing

50
Reed-Solomon Codes

• RS codes are a subset of BCH codes that operate
  at the block level rather than bit level.
  Incoming blocks are represented by a new set of k
  symbols to be packaged in a super-coded block of
  n symbols
• Decoder can detect and correct complete errored
  blocks used in mobile radio / CD as alternative
  / addition to interleaving.

51
Convolutional Coding

• Operates serially on the incoming bit stream and
  the output block of n code digits generated by
  the coder depends not only on the block of k
  input digits but also, because there is memory in
  the coder the previous K input frames. The
  property K is known as the constraint length of
  the code.
• An optimum decoding algorithm, called Viterbi
  decoding, uses a similar procedure.

52
Homework 4

• A mobile data radio link uses interleaving to
  spread the data errors on reception. If the
  interleaving depth is used is a 10 x 8 matrix and
  the bit rate for the signal is 9600 bit/s what is
  the latency introduced by the interleaving
  process?

53
Channel Coding Summary

• Channel coding adds some redundancy, which can be
  used at the receiver to overcome the effects of
  noise or interference introduced in transmission
  over channel.
• Impairments, such as noise, also limit data rate
  that can be achieved.
• Channel Capacity is the theoretical maximum rate
  at which data can be transmitted over a channel,
  under given conditions.
• Advanced schemes such as Turbo coding and LDPC
  approach the capacity.

54
About Channel Capacity

• Impairments, such as noise, limit data rate that
  can be achieved
• For digital data, to what extent do impairments
  limit data rate?
• Channel Bandwidth the amount of frequency
  spectrum we give to each user
• Channel Capacity the maximum rate at which data
  can be transmitted over a given communication
  path, or channel, under given conditions

55
Concepts Related to Channel Capacity

• Data rate - rate at which data can be
  communicated (bps)
• Bandwidth - the bandwidth of the transmitted
  signal as constrained by the transmitter and the
  nature of the transmission medium (Hertz)
• Noise - average level of noise over the
  communications path
• Error rate - rate at which errors occur
• Error transmit 1 and receive 0
• transmit 0 and receive 1

56
Signal-to-Noise Ratio

• Ratio of the power in a signal to the power
  contained in the noise thats present at a
  particular point in the transmission
• Typically measured at a receiver
• Signal-to-noise ratio (SNR, or S/N)
• A high SNR means a high-quality signal, low
  number of required intermediate repeaters
• SNR sets upper bound on achievable data rate

57
Shannon Capacity Formula

• Equation
• Represents theoretical maximum that can be
  achieved
• In practice, only much lower rates achieved
• Formula assumes white noise (thermal noise)
• Impulse noise is not accounted for
• Attenuation distortion or delay distortion not
  accounted for