wn satellite

Document Sample
wn satellite Powered By Docstoc


The origin of satellite communications
Most authorities credit Arthur C. Clarke, famous British science fiction writer
(author of 2001: A Space Odyssey) with originating the idea of a synchronous
communications satellite.
In 1945 Clarke noted that a satellite in a circular equatorial orbit with a radius of
about 42,242 Km (from the center of the earth) would have an angular velocity that
matched the earth, thus it would always remain above the same spot on the ground
and it could receive and relay signals from most of a hemisphere. Three satellites
spaced 120 degrees apart could cover the whole world (with some overlap). Thus,
provided that messages could be relayed between satellites or by a double hop,
reliable communications between any two points in the world would be possible.


  • Clarke had ideas ahead of their times because only on October 1957
    USSR launched SPUTNIK I by using a rocket technology available
    to put a satellite into a low orbit.
  • Synchronous orbit was not achieved until 1963.
  • The SYNCOM series provided the first successful geosynchronous
    communications satellite beginning in 1963, less than 20 years after
    Clarke first conceived the idea.
  • SYNCOM I failed during launch, but SYNCOM II and III were
    successfully placed in orbit on July 26, 1963 and July 19, 1964,
    respectively (NASA & Department of Defence effort).
  • First commercial geostationary satellite: INTELSAT I developed by
    Comsat for Intelsat (1965-1969).
  • Routine operations between USA and Europe began on June 28,
    1965 (birthday of commercial satellite communications).


            Geostationary Earth Orbit SATELLITE

• GEO satellite altitude: 35,786 Km above the equator.
• Revolution around the earth synchronized with the earth’s rotation ==>
  it appears fixed to an observer on the earth’s surface.
• 1 GEO satellite covers about 1/3 of the earth, excluding the high latitude
  areas. 3 GEOs can cover the whole earth surface.
• Orbital period: 24 hours. Thus the geographic area covered by the
  satellite does not vary. Constant use of tracking equipment not required.
• The full-duplex round trip delay through a geo satellite is about 600 ms
  [(A->satellite->B) + (B ->satellite->A)]. In the earlier stage of satellite
  development this delay resulted in some user dissatisfaction. Echo
  cancellers, developed in late 1970 and early 1980 reduced or completely
  eliminated the adverse effects introduced by long round trip delays.


 • Large propagation delay: 1 RTT=240-280 ms, undesirable for real-
   time traffic.
 • FOOTPRINT is the area of coverage of a satellite
 • Cost of the launch is high.
 • Due to high altitude and inherent signal degradation with distance,
   relatively large antennas and transmission power are required for both
   GEO satellites and ground stations.
 • Satellite systems have to share the use of limited spectrum allocations,
   and there is a limit to the number of satellites that can be stationed in a
   given arc of geostationary orbit.
 • Satellites must have sufficient spatial separation to avoid interference.
 • Required spatial separation: between 3 and 6 degrees depending upon
   ES and satellite antenna beam widths, carrier frequency of
   transmission, modulation technique used, and possible performance
   degradation due to interference.


               Medium Earth Orbit SATELLITES

• MEO’s distance from the earth’s surface:
   Ì around 10,000 Km
• MEO round trip propagation delay: 110-130 ms.
• They avoid the large signal attenuation and delay of
  GEO orbits and still allow a global coverage with few
  satellites (10-15 satellites).


               Low Earth Orbit SATELLITES

u LEO’s distance from the earth’s surface:
   F 500-2000 Km Thus, 50-200 satellites are required, depending on
     the degree to which the orbits are controlled
   F lower (500 km) bound dictated by the atmospheric drag which, for
     lower altitudes, reduces the satellite lifetime
   F upper bound (2000 Km) due to the Van Allen radiation belt, which
     entails protection of the on-board equipment against excessive
u LEO round trip propagation delay: 20-25 ms, comparable
  to that of certain terrestrial links.
• Orbit tends to be highly inclined approaching polar orbits,
  so that satellites move rapidly over lands in the North to
  South direction in their about one hour and half orbits.


• Since MEO & LEO are closer than GEO to the earth’s
  surface, the antenna size and the transmission power level
  are generally smaller.
• Footprints too are smaller.
• A constellation of a large number of satellites is necessary
  for global coverage.
• The lower the orbit altitude, the greater the number of
  satellites required.
• As satellites travel at high speeds relative to the earth’s
  surface, a user connection may need to be handed-off from
  satellite to satellite, as they pass rapidly overhead.
• Steerable antennas are crucial to maintain continuous



     •    number of satellites
     •    number of orbital planes
     •    inclination of the orbital planes
     •    relative spacing of the orbital planes
     •    number of satellites in each orbital plane
     •    relative phasing of the satellites of the same orbital plane
     •    relative phasing of the satellites of adjacent orbital planes
     •    orbital height of each satellite and its corresponding orbital


           ADVANTAGES                                LEO               DISADVANTAGES
1. Portables can use low power and non-directive antennas 1. Large number of satellites needed
2. Direct portable-to-portable connections                2. Relatively complex satellites:
3. Short propagation delay                                  - on board switching, routing
4. World-wide coverage                                      - multihop operation for long distance calls
5. Allows high elevation-angle operations from portables    - handoff, cross-links
6. More fail-safe due to the number of satellites           - power management during orbit
7. Provides position location                             3. Complex network operations:
                                                            - satellite monitoring, control, replenishment
                                                          4. Individual satellites spend a large % of time over areas
                                                          with little traffic
  1. Wide area coverage from one satellite                 1. Path delay
  2. Coverage can be directed at traffic concentrations    2. Present expense of Ka band terminals
  3. Cost-effective for limited area coverage              3. Antenna directivity needed at portable
  4. Relatively few satellites needed for world-wide       4. Outages due to rain fades
     coverage                                              5. Portable operation unproven at Ka band
                                                           6. Requires clear line-of-sigh path to satellite
                                                           7. Extended network needed for global coverage


              VSAT SYSTEMS
  • Very Small Aperture Terminals are small, software-
    driven earth stations (typically 0.9-up to 1.5 meters) used
    for transmission of data, video, or voice via satellite.
  • VSAT services are delivered through the use of either C-
    Band or Ku-Band geostationary satellites for video, voice,
    fax and data transmissions.
  • Typically, VSATs use a star network with the use of
    satellite earth stations that rely on a large central hub.
  • Alternatively the use of mesh (hubless) VSAT networks
    can provide communication between VSAT terminals
    directly. They can be configured in both one-way (receive
    only) and two-way (interactive) VSAT terminals.



• Responsible for communication functions.
• Once the satellite is launched, it is very expensive
  and almost impossible to be repaired or upgraded.
• It is required to be simple and robust as the space
  environment with radiation, space debris, etc. is
  harsh for satellites.
• GEOs traditionally serve as bent pipes, i.e. they act
  as repeaters between two points on the earth. In this
  case there is no OBP (on-board processing).


 • Some satellite systems allow OBP, including:
     –   demodulation/remodulation
     –   decoding/re-encoding
     –   transponder/beam switching
     –   routing
 to provide more efficient channel utilization.
 • OBP can support inter-satellite links (ISLs)
   connecting 2 satellites within line of sight.
 • Sophisticated constellations with ISLs allow
   connectivity in space without any terrestrial


         FREQUENCY                      BANDS
• Most commonly used:
   ÿ C band (4-8 GHz)
   ÿ Ku band (10-18 GHz)
   ÿ Ka band (18-31 GHz)
• Some satellite systems use C band and thus employ large
  antennas with a minimum diameter of 2-3 m.
• Most satellites use Ku band for broadcasting as well as for
  Internet connections from server to users, with a terrestrial
  return link.
• Ka band potentially offers much higher bandwidth than Ku
  band, and very small antennas can be used.
• Ka band problem: it suffers from rain fade attenuation.
• Technologies for using frequency bands above the Ka band are
  still immature and further investigation is needed.

 • The communications subsystem is the major component of a
   communication satellite payload.
 • It is usually composed by:
    – one or more antennas, which receive and transmit over
       wide bandwidths microwave signals
    – a set of receivers and transmitters that amplify and
       retransmit the incoming signals.
 • The receiver-transmitter units are known as transponders.
 • Signals transmitted to/from a satellite are known as carriers.




                   Up-link             carrier

                   RF   Low Noise             RF
                        Amplifier (LNA)            Converter   (it is used a selected
                                                               form of digital

Earth station                            70-140- MGz   IF      modulation, typically

                                Base Band
                              burst of bits        MODEM
                              + clock

• “Multiple access” is the technique to share the
  available capacity of a satellite transponder among
  several earth stations which contend for accessing
• The sharing technique may be achieved:
   ÿ by sharing the transponder’s bandwidth in separate
     frequency slots (FDMA)
   ÿ by sharing the transponder’s availability in discrete
     time slots (TDMA)
   ÿ by allowing coded signals to overlap in time and
     frequency (CDMA or spread spectrum). Each earth
     station then separates the signals by recognizing which
     of the codes is destined for itself.


• A communication system must be designed to meet
  certain minimum performance standards, within
  limitations of transmitted power and bandwidth.
• The most important performance criterion is the signal-
  to-noise ratio (S/N if analog, BER if digital) in the
  information channel, which carries the signal in the
  form in which it is delivered to the user.
• S/N (or BER) in a baseband channel depends on a
  number of factors. The most important are:
   ËThe carrier to noise ratio (C/N) of the Radio Frequency
    (RF) or Intermediate Frequency (IF) signal in the
   ËThe type of modulation used to impress the baseband
    signal onto the carrier


                   DEMODULATOR                       S/N analog
                                                      BER digital
FM, PSK, etc.


• An isotropic source , i.e. a transmitting source such as an antenna, in
  free space, radiates a total power Pt W uniformly in all directions. It is
  an ideal source, because it cannot create transverse polarized
  electromagnetic waves.
• At a distance R from the hypothetical isotropic source, the flux density
  crossing the surface of a sphere, radius R, is:

                 F=      W / m2
                    4pR2                                   (1)
• Directive antennas are used to constrain transmitted power to be
  radiated primarily in one direction, thus increasing the power
• The antenna gain is measured in the direction f in which maximum
  power is radiated (boresight of the antenna).
• Gain of antenna, G(f), is a measure of the increase in power radiated by
  the antenna over that from an isotropic radiator emitting the same total


• For a transmitter (satellite or ES) with output power Pt W using a lossless
  antenna with gain Gt the flux density in the direction of the antenna
  boresight at a distance R m (distance travelled in space) is:

          Pt Gt
      F=                           W / m2              (2)
         4p R2
• Pt Gt is also called EIRP (effective isotropically radiated power).
  It describes the characteristics of transmitter and antenna in terms
  of an equivalent isotropic source with power Pt Gt W, radiating
  uniformly in all directions.
• An ideal receiving antenna with an aperture area of A m2 would
  collect a power Pr W given by

                         Pr = FA                 (3)


• A real antenna with a physical aperture area of Ar m2 will not deliver
  this power because:
   – some of the energy incident on the aperture is reflected away from
       the antenna;
   – some of the energy is absorbed by lossy components.
• Reduction in efficiency is described by using an effective aperture

                      Ae = h A r

where h is the aperture efficiency of the antenna.
Ë h accounts for all the losses between the incident wavefront and the
  antenna output port (antenna illumination efficiency, losses due to the
  spillover, blockage, phase errors, diffraction effects, polarization and
  mismatched losses).


• For parabolic reflector antennas:
   *h is in the range 50% to 75%, lower for small antennas
     and higher for large Cassegrain antennas.
• Horn antennas can have efficiencies approaching 90%.
• Thus, the power received by a real antenna with a physical
  receiving area Ar and effective aperture area Ae m2 is

                        PtGt Ae EIRP Ae
   Pr = FAe =                 2 =                                 (4)
                         4pR      4pR 2
• The power received is:
   – independent of the frequency within a given band
   – depends only on
         • the transmitted EIRP
         • the effective area of the antenna
         • the distance R between the satellite and the ES.

• Te GAIN and the AREA OF AN ANTENNA are related by:

      4pAe                       l is the wavelength
   G = 2                         l=light speed/frequency

       l                              Remind that   Ae = h Ar

• Substituting for Ae (m2) in previous Pr formula, we have the Friis
  transmission equation, essential in the calculation of power received in
  any radio link.

                                                         2      Gr is the

               È l ˘                                            gain of the

  Pr = EIRP Gr                                                  receiving

               Î 4p R ˚

         Ae = h Ar

                       EIRP x Receiving antenna gain
  Power received =
                         Path loss = [4pR/ l]2 = Lp
 • In communications systems, dB is used to
   simplify expressions. In dB we have:

         Pr = (EIRP + Gr - Lp ) dBW
         EIRP = 10 log10 (Pt Gt ) dBW
         Gr = 10 log10 (4pAe / l ) dB
         Lp = 20log 10 (4pR / l) dB


 • Previous expression for power received is an ideal case,
   with no additional losses in the link.
 • In practice other factors must be considered in the system
   margin, for computing the link power budget, such as:
    – losses due to attenuation by rain,
    – losses in antenna at each end of the link
    – losses of gain due to antenna mispointing

Pr = EIRP + Gr - Lp - La - Lta - Lra dBW
La = attenuation in atmosphere
Lta = losses associated with transmitting antenna
Lra = losses associated with receiving antenna



• Noise temperature: provides a way of determining how much
  thermal noise is generated by active and passive devices in a
  receiving system. At microwave frequencies all objects, with
  physical temperature Tp > 0 K, generate electrical noise at the
  receiver frequency.
• The noise power is: Pn = k Tn B = No B, where:
   – k is the Boltzmann’s constant = 1.38 x 10-23 J/K =
   – Tn noise temperature of source (space) in Kelvins
   – B bandwidth of power measurement device in hertz (it

     should be the equivalent noise bandwidth)
   – No noise power spectral density (in watts per hertz). The
     density is constant for all radio frequencies up to 300 GHz.


   • In satellite communications we always work with weak
     signals because of the large distances involved.
       =============> thus…………..
   • Noise in the receivers must be reduced as much as possible
     to maintain the best possible carrier-to-noise ratio (C/N)
     and hence the highest quality of communication.
   • This is done by setting the bandwidth in the receiver (usually
     in the IF amplifier stages) just large enough to allow the
     signal (carrier and sidebands) to pass unrestricted, while
     keeping the noise power to the lowest value possible.
   • Noise temperature must be kept as low as possible:
      – immersing the front end amplifier of the receiver in liquid
        helium to hold its physical temperature around 4 K


• Noise temperature from 70 to 200 K can be achieved without
  physical cooling if GaAsFET amplifiers or uncooled
  parametric amplifiers are employed.
• To determine the performance of a receiving system
  (computationally) we need to be able to find the total
  thermal noise against which the signal must be demodulated.
  We do this by determining the system noise temperature Ts.
• Ts is the noise temperature of a noise source, located at
  the input of a noiseless receiver, which gives the same
  noise power at the original receiver, measured at the
  output of the receiver.
• The equivalent noise source Ts is usually located at the
  input of the receiver, replacing the antenna.


• If the overall RF and IF gain of the receiver is
  G (total system gain) and its narrowest
  bandwidth is B, the noise power at the
  demodulator input (i.e. at the output of the
  receiver amplifier) :
   FN = k Ts B Ga = Pn
       • where Ga is the gain of the receiver amplifier
       • Ts is the space + amplifier noise temperature
         (Ts > Tn)


 • If Pr is the signal power delivered by the antenna to the
   receiver at the input to the RF section, the signal power at
   the demodulator input is PrGa representing the power
   contained in the carrier and side bands after amplification
   and frequency conversion within the receiver.
 • The carrier-to-noise ratio at the demodulator input is:

              C   PrGa   Pr
                =      =
              N kTs BGa kTs B
 • It is useful to replace several source of noise in the receiver
   by a single system noise temperature Ts.


 • C/N can also be expressed in terms of receiving
   antenna gain over system noise temperature
   (Gr/Ts) , which is the ratio that can be used to
   specify the quality of an earth station, since
   increasing Gr/Ts the C/N increases as well.
 • Gr/Ts is usually shortened to G/T ratio (called
   figure of merit)
                                           2                                  2
      PG G È l ˘        EIRP È l ˘ Gr
C/ N = t t r          =
       kTs B Î 4p R ˚    kB Î 4p R ˚ Ts
C/N is referred to the down-link.
Ts depends on sky noise temperature, which increases as the elevation angle is reduced
below 10o.

• The down-link of any satellite communication system
  must be designed with the following objectives:
   ‚ to guarantee continuity of the link for a specified %
     of time (typically 99.9%) with a given S/N or BER.
          • it requires a minimum C/N at the receiver input for 99.9% of
            time and probably a modulation scheme or signal-processing
            technique that gives an S/N improvement over the receiver
          • In digital satellite links, phase shift keying (PSK) is the
            modulation method most often used.
•   In order to maintain satisfactory communications, the C/N must remain above
    threshold under all conditions. For an FM system, the threshold is in the range
    4 dB to 15 dB depending on the type of demodulator used. For PSK systems, it
    is typically 8 dB to 15 dB.
•   An allowance, called the system margin, of 7dB is usually used in the Intelsat
    systems for propagation and equipment degradation.


     ƒto carry the maximum number of channels
      at a minimum capital and maintenance cost.
          • compromises between antenna cost, receiver cost,
            tracking accuracy, station manning, modulation, and
            multiple access techniques.
     – We cannot increase the C/N by decreasing the bandwidth in order
       to decrease the noise power unless we are prepared to reduce the
       total number of channels transmitted through the transponder.
     – If we reduce the bandwidth of our carrier, we must use only part of
       the transponder output power so that the remaining bandwidth can
       be used by another carrier. The carrier power C must be shared in
       proportion to the bandwidth used by each carrier and C/N remains
       constant for all carriers


• Design is rather easier than the down-link in most cases,
  since an accurately specified carrier power density must be
  presented at the satellite transponder and it is feasible to
  use much higher power transmitters at ESs than can be
  used on a satellite in most cases.
• Cost of transmitters tends to be high compared to the cost
  of receiving equipment.
• Smaller antenna gain requires greater transmitting power
  for a given EIRP, and the use of TDMA requires still more
  power if the satellite transponder is to be driven into
  saturation. This has the disadvantage that the interference
  level at adjacent satellites rises, since the smaller ES
  antenna inevitably has a wider beam.

• At frequencies above 10 GHz, propagation
  disturbances in the form of fading due to rain
  cause variation in received power level at the
• It may be necessary for each station (or for a
  central station) to monitor up-link attenuation by
  measurement of carrier output power from the
  satellite, so that an increase in the transmitter
  power can be made to compensate for the fade


               (LINK BUDGET)

• The carrier power level at a receiving ES depends on the
  EIRP at the satellite, the ES antenna gain, and the path
  loss, but it is reduced by any losses that occur along the
• For frequencies above 10 GHz rain causes attenuation of
  the carrier, and the attenuation rises rapidly with
• In both analog and digital radio systems there is an
  almost linear relationship between the S/N ratio (or the
  BER) at the demodulator output and the C/N at the
  demodulator input, provided that C/N is above a certain

• Performance objectives of the satellite link: must be
  specified either in terms of minimum allowable S/N (or
  BER for a digital signal), or as a minimum allowable (not
  lower than a certain value) C/N, which is generally
  specified for a percentage of time.
• Factors that contribute to noise in an ES receiving channel
   –   the receiver thermal noise
   –   losses in waveguides and waveguide components
   –   sky noise
   –   interference entering the ES receiving antenna
   –   interference entering the satellite receiving antenna
   –   thermal noise generated in the satellite
   –   intermodulation noise throughout the system
   –   many other sources of noise may exist in a particular system


  • Link budget allows to compute the C/N at the
    demodulator input.
  • It takes into account the ES and satellite
    characteristics in terms of EIRP, G/T, all power
    losses, thermal and interference noise.

            1        1          1         1
                =         +           +
           C / N (C / N)up (C / N)down C / Itot

where the quantities are not expressed in dB.
Itot is the total interference power.
(C/N)up is computed analogously to the (C/N)down, by
considering the ES EIRP and the satellite G/T.
The noise or interference powers can be added as incorrelated among them.


  • Coding techniques are applicable to satellite transmissions
    in order to improve the system performance.
  • There are two basic alternatives to be explored when
    coding is considered: block coding and convolutional
  • Before to treat them, we must introduce the concept of
    channel capacity and the Shannon bound.


We made the assumption that data are generated by a memory-less source. The output
   of the source can be viewed as a time sequence of source states
          A=(a1, a2, a3, ……, aj, ……)
   with each of these states chosen, according to some probability rule, from a finite
   source alphabet X,
          X = (x1, x2, x3, ……xn).
As the simplest example, a binary source has a 2-element alphabet, defined by X=(0,1),
   and by a typical output such as A=(1, 1, 0,0,0,0,1,1,0,……)

Shannon, in his landmark paper (1948) defined the information
  associated with a source output (or, in general, associated with any
  event), to be the negative logarithm of the probability of that event.
  Thus, the information associated with a source state xj is
                 I(xj)= - loga p(xj)
Where p(xi) is the probability of source state xi.

 • Base of the logarithm is arbitrary: simply determines the
   units in which the information is measured. The base a is
   universally taken as 2, thus the corresponding measure of
   information is the bit.
 • What does the previous formula indicate?
      – If the probability of an event is very high, then p(xj) is
        approximately 1 and the log of 1 is zero. It means that little
        information is associated with an almost certain event.
      – If the probability of an event is very small, its negative
        logarithm is large, thus indicating a great amount of
 • In typical applications it is not the information associated
   with the source states that is important, but rather the
   average information associated with the source. This
   average information (averaged over all source states) is
   referred to as entropy.


  • Shannon showed that if a source produces
    information at a fixed rate R and the available
    channel has capacity H, then a coding and
    modulation scheme exist such that the decoded
    error probability can be made arbitrarily small for
    R<H. For R>H, the application of coding usually
    only further degrades system performance.


• For an additive White Gaussian noise channel, the information
  capacity H of the channel is given by:

                              Pr                    Shannon-Hartley
 H = B log 2 (1 +                 ) bps             law
                             No B
B = channel signal bandwidth (in hertz)
Pr = received signal power (in watts)
No = noise power spectral density (in watt per hertz).
No B is the noise power, usually indicated by N


• We can rewrite the Shannon law specifically for a digital
  communication link by putting the bit duration Tb = 1/H
  (thus H = 1/ Tb)
• Energy per bit is Eb = Pr Tb = Pr / H ==> Pr = Eb H
• Substituting P in previous formula:

           H             E H
             = log 2 (1 + b   )
           B             N0 B
  which illustrates the relationship between Eb/N0 and H /B.
  H /B is called spectral efficiency of the communication
  link, i.e. the ratio of the bit rate to the bandwidth of the


  H < B for operation at capacity (power limited link because it does not
  use its its bandwidth efficiently).

• Regardless of the bandwidth used, Eb/N0 approaches to a finite
  limit as B -> infinite. This limit is:
  Eb / N0 = log e 2 = - 1.6 dB                               bound
• This value sets a lower theoretical limit on the Eb/No we can
  use in any (error free) communication link, regardless of the
  communication or the coding schemes.
• This value indicates that reliable transmission, with as low
  error probability as desired, is possible on the AWGN channel
  only with a ratio of Eb/N0 at least this value.



• Theoretically it is possible to operate a link at Eb/No= -1.6
  dB; in reality available demodulators limit the minimum
  channel bit energy (= data bit energy x coding rate) to
  around Ec/No = 0dB.
• Below the level the demodulator may not be able to
  recover a reliable bit timing clock and will not sample
• Power efficiency may be expressed:
   – in terms of the C/N required in order to have an acceptable
   – in terms of the required average received bit energy-to-noise
     density ratio for a given BER (i.e, BER =f(Eb / No)).
• In practical measurements, it is more convenient to
  measure the average C/N because bit energy meters are not
  commonly manufactured.


  • The following relations are useful for the Eb/No to
    C/N transformation:

Eb/No equals the product                        Ê1ˆ
                                 Eb = C Tb = C Á ˜
of the C/N ratio and of the                     Ë H¯
receiver noise bandwidth-to-            N
                                 N0 =
bit rate ratio                          B
                                 Eb    CTb   C / H CB
in BPSK: Eb/No =C/N                  =     =       =
                                 N0 N / B N / B NH
in QPSK: Eb/No =1/2 C/N
                                 Eb C B
                                 N0 N H


  • BER of digital signals can be improved by the use of
    error detection techniques, that add redundant bits to
    a data stream in such a way that an error in the data
    stream can be detected (error detection codes) and
    corrected (forward error correction, FEC).
  • Unless one redundant bit is added for every data bit,
    the exact position of a single bit error cannot be
  • Usually one redundant bit is added for every N data
    bits, thus allowing a single error within that block of
    N bits to be detected.

   ß 8-bit ASCII coding
   ß Linear block codes
   ß Binary cyclic codes
      vBCH (Bose-Chaudhuri-Hocquenghem) codes
      vGolay codes
ÿ Convolutional codes
ÿ Punctured codes
ÿ Turbo codes


• Sometimes called VITERBI codes because the
  algorithm for decoding was studied by Viterbi.
• Are generated by a tapped shift register and two or more
  modulo-2 adders wired in a feedback network.
• The name is given because the output is the convolution of
  the incoming bit stream and the bit stream that represents
  the impulse response of the shift register and its feedback
• Each incoming information bit propagates through the shift
  register and influences several outgoing bits, spreading the
  information content of each data bit among several
  adjacent bits.

• Selected bits are erased before transmission according
  to a set puncturing pattern, known to both the coder
  and the decoder.
• Code rate increased. Example: a 1/2 rate code may be
  punctured to a 7/8 code by removing 3 bits out of
  every 7.
• Gain difference between a punctured code and a non-
  punctured one of the same rate may be made as small
  as 0.1 dB by appropriate choice of code and
  puncturing pattern.


             TURBO CODES
• A turbo encoder is a combination of two simple encoders.
• The input is a block of K information bits sent uncoded.
• The two encoders generate parity symbols from two simple
  recursive convolutional codes, each with a small number
  of states.
• Key innovation of turbo codes is an interleaver P which
  permutes the original K information bits before input to the
  second encoder.
• The permutation P allows that input sequences for which
  one encoder produces low-weight (few non-zeroes)
  codewords will usually cause the other encoder to produce
  high-weight (many non-zeroes) codewords.

• Even though the constituent codes are individually
  weak, the combination is surprisingly powerful.
• The resulting code has features similar to a
  “random” Hock code with K information bits.
• Random Hock codes are known to achieve the
  Shannon limit performance as K gets large but at
  the price of a prohibitively complex decoding
• Turbo codes mimic the good performance of
  random codes, for large k, using an iterative
  decoding algorithm based on simple decoders
  individually matched to the simple constituent

• Multiple access is the “ability of a large number of earth
  stations to simultaneously interconnect their respective
  voice, data, teletype, facsimile, and television links
  through a satellite”
• The basic problem involved is how to permit a changing
  group of earth stations to share a satellite in a way that
  optimizes the
   –   satellite capacity     - adaptability to multimedia traffic
   –   spectrum utilization   - cost
   –   satellite power        - user acceptability
   –   interconnectivity      - flexibility
• There are three basic techniques: FDMA, TDMA, CDMA


• In all the three classical multiple access techniques
  some resource is shared.
• If the proportion allocated to each earth station is
  fixed in advance, the system is called fixed access
  (FA) or pre-assigned access (PA).
• If the resource is allocated as needed in response
  to changing traffic conditions, the multiple access
  system is termed demand assignment (DA).


              TRAFFIC TYPES
 • Real-time (stream)
    – some errors accepted
    – delay requirements very restrictive
    – ex: CBR, VBR, voice,video..
 • Non real-time
    – no errors
    – delay requirements more flexible
    – ex. file transfer, ............
 • Interactive
    – some errors and delay acceptable
    – ex. messaging, internet enquiry,.....

• An earth station is permanently assigned a carrier
  frequency (or several) and a bandwidth around that
  carrier. The station modulates all of its outgoing
  traffic (whatever the destination) on that carrier.
• An originating station’s traffic capacity is limited by
  its allocated bandwidth and the C/N that it can
  achieve on the down-links.
• The carrier frequencies and bandwidths assigned to
  all the earth stations constitute a satellite’s frequency


• In TDMA a number of earth stations take turns transmitting
  bursts through a common transponder.
• A group of earth stations, each at different distance from the
  satellite, must transmit individual bursts of data in such a way
  that bursts arrive at the satellite in a prescribed order.
• Stations have to adjust their transmissions to compensate for
  variations in satellite movements, and they must be able to
  enter and to leave the network without disrupting its operation.
• These goals are accomplished by organizing TDMA
  transmissions into frames containing reference bursts that
  establish absolute time for the network.
• Each frame contains a reference burst and a series of traffic


• Digital data streams from many sources are transmitted
  sequentially in assigned time slots.
• The time slots are organized into frames that also allow the
O Difference with TDM:
   – while in TDM every data comes from the same transmitter
      and the clock and time frequencies do not change, in
      TDMA each frame contains a number of independent
• Each station has to know when to transmit and it must be able
  to recover the carrier and clock for each received burst in time
  to sort out all wanted baseband channels.
• This is not an easy task to perform at low C/N ratios. A long
  preamble is generally needed which decreases system

• An earth station may transmit into or receive from
  several transponders (transponder hopping).
• Reference bursts are generated from a control station
  on the ground (master station) in a centralised
  control satellite network.
• A reference burst contains at least the following
   Fcarrier and bit timing recovery sequence (CBTRS) for
    synchronizing the transmitting and receiving carriers and
   Fa unique word (UW).
   Fthe station identification
   Fthe network housekeeping information


   • Each burst starts with a preamble, which provides
     synchronization and signalling information and
     identifies the transmitting station.
   • Reference burst and preambles constitute the frame
   • The smaller the overhead, the more efficient a
     working TDMA system is, but the more difficulty
     it may have in acquiring and maintaining
   • The minimum frame length is 125 ms required by
     a voice channel sampled at 8-kHz rate. The
     maximum is arbitrary.


                  TDMA FRAME STRUCTURE

             1         2       3       4 … n-1         n

                                                  Traffic burst
       Guard time                                 from station 4

Reference burst
                   CBTRS+UW        traffic bits



• CBTRS consists of a short transmission of unmodulated
  carrier followed by carrier phase transitions between 0 and
  p radiants at the symbol clock frequency.
• The RB UW serves to mark the beginning of valid data. At
  the receiving end of a link incoming bits are clocked into a
  shift register where they are compared with a stored version
  of the expected UW (UW correlator). When the bits in the
  register match the stored version, the correlator produces its
  maximum output voltage pulse.
• UW sequencies and thresholds are chosen to minimize the
  false alarm and maximize the UW detection.


• Centralized case:
    – generally more robust because one station alone is responsible for
      the BTP consistency.
• Distributed case:
    – as all stations have to listen to each other, the chance that some
      control information is lost or misinterpreted and the BTP corrupted
      is increased.
    – more responsive to traffic variations, because only one RTT is
      needed for the network, so highly bursty traffic sources are more
      efficiently dealt with, thus increasing the global channel
    – lack of robustness may be compensated with complex recovery
      algorithms (some additional channel overhead necessary).

• On the average, a number of users occupy all of a transponder
  bandwidth all of the time.
• The signals are encoded so that information from an individual
  transmitter can be detected and recovered only by a properly
  synchronized receiving station that knows the code being used
* Decentralized satellite network: only the pairs of stations that are
  communicating need to coordinate their transmissions.
* No frequency (as in FDMA) or time slot (as in TDMA)
  coordination with any central authority is required.
* Suited for military tactical communications environment, where
  many small groups of mobile stations communicate briefly at
  irregular intervals.


   • CDMA is the application of a spread spectrum technique
     (which is the inverse of the bit packing technique, where as
     many bits as possible are packed in 1Hz of bandwidth).
   • One bit is spread over some or many hertz.
   • 2 types of spread spectrum: direct sequence (DS, one bit is
     spread in phase), or frequency hop (FH, one bit is spread in
   • In FH CDMA, multiple users can transmit on the same
     frequency with only some minimal interference one to
     another, due to the powerful antijam properties of spread


• At frequencies above 10 GHz rain is the dominant
  factor in satellite transmitting signal propagation.
  Rain propagation has been intensively studied since
  the late 1960s [1].
• Rainfall rate rather than the accumulation determines
  propagation effects.
• Long-term behavior of rain is described by a
  cumulative probability distribution (or exceedance
  curve) that gives the % of time ( % of one year) that
  rain rate exceeds a given value.


• If the rain rate R is constant over a path of L Km,
  the attenuation A caused by rain is given by:
          A= a RbL         dB
• a Rb is called specific attenuation
• coefficients a and b strongly depend on frequency.
  Tables give their values.
• Usual situation for terrestrial radio links.
• On satellite paths rain rate changes with position
  of the stations and with time.
• Also the length L of the portion of path that
  contains rain is also variable (L(t))


  • The estimated attenuation to be exceeded for a
    percentage p of an average year, in the range
    0.001-1% (Ap), in terms of percentage of time in a
    year is given by the ITU-R interpolation formula,
    when the attenuation exceeded for 0.01% of a
    year (A0.01) is known for a given site.
                  Ap = A0.01 0.12 p-( 0. 546+0 .043log   p)

The value p corresponding to a given value of Ap may be computed
by inverting the above equation:
                11.628 ( -0.546+ 0.298 + 0.172 log 0.12 A0 .01 / A p
   p = 10
                with the constraint that A0.01/Ap≥0.15

  • Fading at the Ka band over 99% of the year is less
    than 2 dB for most locations within Europe.
  • It is required a technique to increase the availability
    of a link when needed, without resorting to a large
    fixed fade margin (fade countermeasures).
  • Possible techniques:
      –   (frequency, site) diversity              -adaptive coding
      –   adaptive transmission rate               -fade spreading
      –   up-link power control                    -adaptive TDMA
      –   beam shaping                             -land lines
      –   alternative satellites

• Storms are localized events, with heavy rainfall
  occurring over a small raincell of area of few Km2.
• As storms occur only in the lower atmosphere, earth
  stations separated by a distance greater than that
  covered by each raincell experience statistically
  independent fading.
• Minimum separation for statistical independence is
  F(raincell size, satellite elevation).Typically 10 Km.
• In spatial diversity, a link shares 2 ESs; terminal with
  best reception is selected.
• “Diversity” gain may reach over 10 dB at 30 GHz.

  • Resource pool:
     – additional bandwidth at a lower RF
       frequency less affected by fading (from
       Ka band to Ku band).
  • When the margin on a particular Ka band
    link becomes unable to overcome rain
    attenuation, the link is switched to a lower
    frequency band through a frequency

  • Frequency scaling formula (in the frequency range 7-50
    GHz): gives the attenuation A2 (over the average year) as a
    function of frequency and attenuation A1 :

                  A2 = A1 (j 2 / j 1 )1- H (j 1 ,j 2 , A1 )
  • where                   f2
             j(f) =
                       1 + 10-4 f 2
             H(j1 ,j 2 , A1 ) = 1.12 *10 -3 (j2 / j 1 )0.5 (j1 A1 )0.55
  • and A1 and A2 are the equiprobable values of the excess
    rain attenuation at frequencies f1 and f2 (GHz),
  • In the [20-30] GHz band, attenuation at 30 GHz is about 2
    times than attenuation at 20 GHz.


            ADAPTIVE TDMA

• If multiple access scheme is employed, an alternative
  form of diversity, known as “ common resource
  diversity” may be used.
• Additional transmission time from a common resource
  pool is offered to links that are experiencing fading.
• The extra time increases the average power of the link,
  thereby overcoming the fade, and may be used for error
  correction coding.


 • Resource pool:
     – additional bandwidth at a lower RF frequency less
       affected by fading
     – spare time-slots in a TDMA frame
 • As only a few stations will generally suffer fading
   at the same time due to stations’ spatial
   incorrelation, the common resource pool requires
   only enough capacity to overcome those fades.


   Adaptive Forward Error Correction
• If capacity of the satellite link is permitted to decrease
  during a fade, then FEC coding schemes con be used
  as fade countermeasure.
• Adaptive coding may be used only on those links that
  can tolerate a reduced throughput when faded.
• During a fade the data rate is reduced and extra
  coding information is inserted into the channel so that
  the channel rate remains constant.
• Gain realized by the coding schemes: 2-10 dB
  depending on the scheme adopted and operating

• Coding rate may be adapted to the channel
• A 1/2 code reduces the data rate by half providing
  an additional fade tolerance of 3 dB.
• When a reduced throughput is not tolerated by the
  faded link, coding application needs a wider
  bandwidth, which can be obtained from the
  common resource pool.


• A simple method is to restore the bit energy/noise energy
  ratio to the unfaded value.
• Done by
   – increasing the transmitted power
   – by reducing the data rate
• Fade spreading technique works by decreasing the data
• Simple rate reduction increases the bit energy, and the
  power per unit bandwidth is also increased.
• This may lead to interference (intermodulation) and related
  problems, possibly requiring the application of energy

• With fade spreading it is possible to reduce the data
  transmission rate without reducing the bandwidth,
  while still benefiting from an increased bit energy.
• Data stream is combined with a pseudorandom code
  (PRC), known as spreading function, generated from
  a maximal-length tapped shift register tailored to the
  required bandwidth.
• The combination process spreads the data signal to
  the bandwidth of the PRC.
• Spread spectrum signal transmitted to the receiver
  that recombines it in the inverse function with the
  same PRC and a low-pass filter de-spreads the
• Improvement in the data signal-to-noise ratio.         81

• If an ES is equipped with up-link power control,
  fades on the up-link to a satellite are compensated
  for by an increase in transmitter output power.
• Use of UPC implies that the ES is not operating at
  the maximum efficiency, thus making the link
  more sensitive to down-link fades.
• If a fade of 10 dB occurs on the up-link and the
  up-link transmitter increases power by 10 dB to
  compensate for it, the down-link of the receiving
  station will be unaffected because the transponder
  output toward it will be the same.

  • In FDMA, if all carriers using the transponder work in
    an intermodulation-free environment, a small amount
    of down-link power control is available if additional
    power is used on the up-link toward a faded down-link
    (saturation problem must taken into account).
  • Additional down-link power is obtained by increasing
    the up-link power of the faded carriers
     ‡attention: not to reduce the transponder back-off to
       limit intermodulation!
  L Disadvantage: under unfaded conditions the quality of the
    up-link is worse than that of the fixed power scheme for
    the same maximum power output by the dynamic range of
    the power control scheme.
  • The basic concept is the same as to reserve a pool of
    resources for faded links


• The rate at which data are transmitted is varied
  according to propagation conditions.
• Higher data rates are used in unfaded conditions,
  and progressively reduced as fading occurs.
• If TX bandwidth is reduced, the RX must reduce
  its bandwidth to that of the data signal to obtain
  maximum gain by rejecting noise out-of-band.
L DISADVANTAGE: a variable bandwidth filter in
  the receiver which, added to a variable rate
  demodulator, increases the system’s cost.

 [1] Louis J. Ippolito, R.D. Kaul, R.G. Wallace
     “Propagation effects handbook for satellite system design”, NASA Reference
     Publication 1082(03), Washington, DC, 1983.
 [2] Kamilo Feher
     “Digital Communications. Satellite/earth station engineering” Prentice-Hall.
 [3] Timothy Pratt, Charles W. Bostian
     “Satellite Communications.” John Wiley & Sons.
 [4] N. Celandroni, E. Ferro, N. James, F. Potorti'
     "FODA/IBEA: a flexible fade countermeasure system in user
     oriented networks", International Journal of Satellite Communications,
     Vol. 10, N. 6, pp. 309-323, November-December 1992.
 [5] N. Celandroni, E. Ferro, F. Potorti'
     "A traffic generator for testing communication systems:
     presentation, implementation and performance", Real-Time Systems,
     Vol. 13, No. 1, pp. 5-24, July 1997.


[6] Gerard Maral, Jean-Jacques de Ridder, Barry Evans, M. Richharia
    “Low orbit satellite systems for communications”, International Journal of
    Satellite Communications, Vol. 9, N. 6, pp. 209-22, 1991.
[7] Gerard Maral
    “VSAT networks”, Wiley 1995.
[8] Gerard Maral, Michel Bousquet
    “Satellite communication systems: Techniques and technology”, Wiley
[9] R. Peterson, R. Ziemer, D. Borth
    “Introduction to spread spectrum communications”, Prentice Hall.
satellite communication references:
Books on satellite communications:



Shared By: