When the computing function is apportioned among cpus spread geographically and connected by a communications systems it is called?

II Basic VSAT Network Design Concept

A small earth station is physically limited by a small antenna diameter. A small antenna poses difficult problems in establishing a good satellite link. The small antenna has low antenna gain in the transmit frequencies. The cost and regulatory elements limit the amount of transmitter power a station can possess. The small antenna also has low antenna gain in the receive frequencies, and again, cost constraints limits the ability of the low-noise amplifier (LNA) that can be used in the receiver. Thus, a small earth station has an intrinsically poor receive-noise figure (NF), a high antenna sky-noise floor, and a low transmit equivalent-isotropic radiated power (EIRP) performance. From the satellite link budget analysis point of view between two satellite earth stations, it is clear that these performance deficiencies of a small earth station can be amply compensated by having a second, large earth station as a partner. The large station possesses ability to provide good transmit EIRP to match the poor receive noise figure of the small station, so long as it is within the satellite downlink power flux density constraints. In addition, the good receive noise figure of the large station can be used to offset the low EIRP of the small station. Thus, in expanding this two-station relationship to a network of multiple small earth stations such that the large earth station serves as a central hub station, a natural “star connectivity” network topology evolved. Of course, it is also possible to have two equally capable earth stations communicating with each other in a single-channel-percarrier (SCPC) basis, or multiple equally capable earth stations in a mesh network situation. From the preceding discussion, these variations do not maximize network infrastructure cost and data transport efficiency to the same extent as the “star connectivity” network in many respects and for obvious reasons. In general, the “star connectivity” network topology predominates in the VSAT industry.

Figure 1 shows a pictorial representation of a typical VSAT network in a star-connection configuration. In the continental United States (CONUS), the physical-hub earth-station antenna sizes used in a Ku-band satellite network range from 4 to 10 m in diameter. The Ku-band VSAT antenna sizes typically range from 1.2 to 1.8 m in diameter. In areas where there is marginal satellite footprint coverage, VSAT antennas with 2.4-m diameter are often necessary. The selection of antenna sizes is the result of the satellite link budget calculation. There is an interesting observation from the link budget analysis that is necessary to highlight. It can be said that the feasibility of a VSAT network in terms of satellite link calculation benefits a great deal from the average additional 5 dB of coding gains derived from the powerful modern forward-error-correction (FEC) coding techniques.

An actual commercial VSAT network link budget for the network outroute signal worksheet is shown in Table I. Outroute path is referred to as the signal path from the transmitter of the hub station to the receiver of the VSAT station. The inroute link budget, the signal path from the transmitter of the VSAT to the receiver of the hub station, is shown in Table II. A more in-depth link budget calculation example including independent uplink and down link rain fade considerations can be found in Cheah (1999).

TABLE I. Outroutea Link Budget Calculation Example for a Commercial 800 VSAT Station Network Belonging to a Hotel Chain

TABLE II. Inroute Link Budget Calculation Example for a Commercial 800-VSAT Station Network Belonging to a Hotel Chain

aInroute is defined as the path from the remote VSAT to the hub.

A VSAT network uses well-defined frequency channels within a satellite transponder frequency assignment. The network uses a time-division-multiplex (TDM) technique for the outroute path where data packets are sent continuously by the hub station. There are no intentional breaks in the signal transmission. The inroute path connection is based on the time-division-multiple-access (TDMA) technique. VSAT transmits only when it is required to do so. The network access is typically initiated through the slotted-ALOHA random access protocol, such that a VSAT within the network can contend for the network access privileges. All or selected VSAT stations are told by the hub that within a fixed designated time slot, VSAT can gain network access permission by transmitting a network access request. After the network access permission is granted to a VSAT, its network connection session bandwidth is assigned based on the demand-assigned-multiple-access (DAMA) method. In most networks, multiple frequency-division-multiple-access (FDMA) inroute channels are also available for network connection assignments. Figure 2 shows an actual network spectrum plot as seen by the Kuband geostationary satellite for a 512-kbps outroute signal from the hub station and a 128-kbps inroute signal from the VSAT station. This plot was taken from a standard VSAT video intermediate frequency (IF) port, and since the inroute transmission is intermittent, the spectrum plot was generated using the spectrum analyzer max-hold function.

As the geostationary satellite altitude is about 35,784 km above ground, the skewed signal path from each VSAT scattered on the ground has different time delays. Therefore, all VSATs must keep correct absolute timing as referenced at the hub station. Similarly, the transmit power of each VSAT as seen by the hub station may also be different because of the geographical satellite-coverage footprint variations and path losses. Thus, as far as inroute physical layer transmission is concerned, each VSAT must at least keep a configuration of its exact time delay and power settings with respect to the hub station. Naturally, there is a multitude of other station- and network-specific parameters that a VSAT must have in its station configuration, such as its network identification (ID), VSAT station ID, and the network-connection-related parameters.

VSAT network system design is considered as hub-centric, so that all references to system time, power, and network connections are viewed with respect to that of the hubstation. It is not surprising that a VSAT station is commonly referred to as a “remote.” Thus, the network's single point of failure lies with the hub. A robust and high-performance hub design is therefore imperative to the success of the VSAT network infrastructure construction.

2.1 Scattering of Electromagnetic Waves at Mono- and Polydisperse Particle Ensembles

When observing volcanic eruptions using radar systems, an EM wave is transmitted in form of a narrow, cone-shaped beam (its opening angle depends on wavelength and antenna diameter, and is typically less than a few degrees). This beam interacts with the medium it passes through, and if objects are in the beam, some of the EM energy is scattered. A small fraction of this scattered energy is scattered directly back towards the instrument and can be received by appropriate antennas. Obviously, the amount of energy received back at the instrument must be related in some way to the number and properties of the scatterers in the radar beam. It turns out that the important quantity to determine is the so-called radar cross section, σ (for single targets) and the radar reflectivity, η, or radar reflectivity factor, Z (for distributed targets). Note, all symbols used are summarized in Table 2.

Table 2. List of Symbols

Calculation of the radar cross section requires investigation into the interaction of an EM wave with the scatterers. This problem was first investigated by Mie (1908). The Mie (or sometimes also called Lorenz–Mie) theory (for details, see eg, Stratton (1941); Mishchenko et al. (2002) or the short summary given in the appendix of Gouhier and Donnadieu (2008)) basically fills the gap between what is known as Rayleigh scattering and geometrical optics. In the case of Rayleigh scattering, the electrical field inside the scattering object is considered homogeneous, meaning that all molecules contribute equally (in-phase) to the scattering. Therefore, the scattering amplitude is proportional to the number of molecules, which in turn is proportional to the particle's volume (α D3, D being the particle diameter). The particle's radar cross section corresponds to the electrical field intensity (amplitude squared) and is thus α D6. Conversely, in the case of geometrical optics, scattering is independent of wavelength and scales with the object's diameter squared (α D2), and thus with its true cross section. In between these two regions is the so-called Mie region, where the interaction of the EM wave with the object/particle needs to be considered in detail. In case of Rayleigh scattering, forward and backward scattering (ie, back to the source) equal each other, while in case of Mie scattering, forward and backward scattering differ significantly and are highly dependent on frequency (see eg, Fig. 3 in Gouhier and Donnadieu (2008)). It can be shown that Rayleigh scattering and geometrical optics are special end member cases of the Mie theory.

The radar cross section depends on the particle size or PSD, as well as on the dielectric permittivity of the particles. The effective dielectric permittivity or constant, ε, of volcanic ash has been measured by several authors in different frequency ranges (Campbell and Ulrichs (1969) at 450 MHz and 35 GHz, Adams et al. (1996) between 4 and 19 GHz, Oguchi et al. (2009) between 3 and 13 GHz, and Rogers et al. (2011) between 65 and 110 GHz). The data show that there is only a slight decrease (10–20%) of the dielectric constant with silica content of the material (Adams et al., 1996; Oguchi et al., 2009), and it appears to be almost independent of frequency. Therefore, most of the studies of radar application on volcanoes use the value of ε = 6 + 0.15i, which is an average value of Adams et al. (1996). Instead of reporting the dielectric constant, one often finds the complex dielectric factor, K, that is related to the dielectric constant by

[1]|K|2=| ε−1ε+2|2.

The backscatter cross section is proportional to |K|2 if the particles are spherical and the same size. Converting the dielectric constant ε to K gives K = 0.39 ± 0.02 for volcanic ash (Adams et al., 1996). Comparing K of volcanic ash to the dielectric factor of water (0.93) and ice (0.197), ash particles are 2.4 times less reflective than water and 2 times more reflective than ice (Marzano et al., 2006a).

Water vapor is one of the major volcanic gases, and upon rise of the eruption cloud condensation and ice formation is commonly observed (eg, Durant et al., 2008), especially because ash serves as a nucleation site. Importantly, a thin water film of 10 μm thickness on a particle of 100 μm size brings the ash dielectric factor close to that of water (see Fig. 9 in Oguchi et al., 2009). Therefore, condensation can change the overall reflectivity of an ash cloud. Other factors influencing the dielectric factor are the porosity of the material, which affects the density of the ash particles (see eg, p. 197/8 in Marzano et al., 2012b, for a short discussion).

The second and even more important variable controlling the reflected energy is the PSD. The PSD is a direct consequence of the fragmentation process (see Part 2) and can vary strongly between different eruptions. Several different distributions have been used to describe PSDs (eg, log-normal distribution (eg, Bronstein and Semendjajew, 1980), Gamma distribution (eg, Kotz and Johnson, 1988), Weibull distribution (eg, Brown and Wohletz, 1995), see Fig. 1B); however, there is still no real consensus on which distribution fits different eruption types best (Gouhier and Donnadieu, 2008).

The backscatter cross section, along with the extinction (that is, the sum of scattering and absorption) of a single particle of 2 mm radius as a function of wavelength of the incident wave is shown in Fig. 1A. Rayleigh scattering occurs as long as scattering increases linearly with the so-called size parameter, x, as long as x = 2πr/λ < 0.7 or D = 2r ≤ 0.7λ/π. This is consistent with estimates made for the Rayleigh region by Gouhier and Donnadieu (2008). The region of geometrical optics is reached once scattering is constant with respect to x, ie, x > 50 or D = 2r ≥ 50λ/π. These constraints still hold in case of wider (see Fig. 1C) or broad PSDs (Fig. 1D). The Mie region that was so prominent for a very sharp distribution in Fig. 1A disappears and becomes flat (Fig. 1D).

From those figures, it is obvious that as long as scattering occurs in the Rayleigh region, there is a clear defined relation between the radar cross section and the wavelength of the incident wave. This is advantageous, as the total backscattered energy (see below) can be easily calculated. But this also shows the dilemma of observing volcanic eruptions using radar. In order for large lapilli to even lie in the Rayleigh region, a large wavelength is required (see Tables 3 and 4). However, a large wavelength also means that the cross sections of small particles are too small to be seen by the radar. On the other hand, if we choose a small wavelength to increase the sensitivity (ie, radar cross section) of small particles, the larger ones move into the Mie region, and the simple relation between particle size/wave length and radar cross section breaks down, making the determination of the PSD or even the total amount of ash in the field of view almost impossible. In addition, at higher frequencies (X-band and above) and high particle concentrations, path attenuation becomes more important and has to be considered in detail (see Marzano et al., 2006a, especially Eq. 13 therein, as well as p. 319 and following).

Table 3. Frequency Bands and Wavelengths of Radar Systems and Particle Diameters Below Which Scattering can be Described by Rayleigh Scattering or Above Which by Geometrical Optics (See also text)

The radar systems mentioned have been used to observe volcanic eruptions. Please note that the systems mentioned are only examples, and several other systems have also been used.

Table 4. Chart Associating the Different Scattering Regimes of Different Particle Diameters With the Typical Wavelength Used in Radar Observations

The Letters R, M, G indicate the Rayleigh, Mie, and geometrical optics scattering regimes, respectively. The addition nd means that those particles are most likely not detectable by that radar system if particle concentrations are too low. In the case of blocks, we have listed the sizes above which scattering is properly described by geometrical optics.

The classification of ash is based on Rose, W.I., Bluth, G.J.S., Ernst, G.G.J., 2000. Integrating retrievals of volcanic cloud characteristics from satellite remote sensors: a summary. Philosophical Transactions of the Royal Society of London A: Mathematical, Physical and Engineering Sciences 358 (1770), 1585–1606; Marzano, F., Picciotti, E., Vulpiani, G., Montopoli, M., 2012b. Synthetic signatures of volcanic ash cloud particles from X-band dual-polarization radar. IEEE Transactions on Geoscience and Remote Sensing 50 (1), 193–211.

II Design of a Radio-Interferometric Array

The objective of a radio-interferometric array is to sample the coherence function over as much of the u-ν plane as possible. At most wavelengths, the radiation is collected using parabolic reflectors which focus the radiation to a focal plane; typical antenna diameters are 5–100 m with 25 m being a popular compromise between construction cost and sensitivity. The ability to point the antennas and the surface accuracy limits the highest practical observing frequency. At wavelengths longer than about 50 cm, dipoles become favored because of the large antenna diameter that would be required to produce reasonable antenna gains.

Amplification of the radiation is vital at most wavelengths. Over the years receiver technology has advanced steadily to decrease the noise introduced at this stage and to increase the bandwidths feasible. Current state-of-the-art amplifiers use cryogenically cooled field-effect transistors (FETs) or high-electron-mobility transistors (HEMTs) for centimeter wavelengths, and either Schottky devices or SIS devices at millimeter wavelengths. At meter wavelengths, cryogenically cooled amplifiers are not required because the noise introduced by the sky typically dominates that introduced by the receiver.

Once the signals have been collected and amplified, they are relayed to a central location for estimation of the coherence. Often this is more easily done at low frequencies, so heterodyne techniques are used to mix the signal down to an intermediate frequency (IF) for transmission to the central location. For this conversion an accurate frequency standard is required at each antenna in order to maintain coherence throughout the array. This frequency standard may be either distributed from a central location (via cables, waveguides, radio links, or optical fibers) or derived from an extremely stable clock such as a hydrogen maser for widely separated antennas. The IF signal is then transmitted to the central location; for short distances the same route as the frequency standard can be used, whereas for long distances (>100 km) the signals are commonly digitally sampled and recorded on magnetic tapes which are then shipped to the central location for later processing. Bandwidths allowed by current technology exceed 100 MHz.

At the central location the signals must be multiplied and integrated (i.e., correlated) to form estimates of the coherence function. As digital correlators are preferred almost universally for their lower levels of systematic errors, the signals must be digitized before correlation. One-bit digitization has been popular for many years and leads to only a small loss in the signal-to-noise ratio. The advent of cheap, fast samplers has spurred more elaborate digitization schemes, such as three-level encoding, which usually are preferred for their lower loss of signal to noise. Before the correlation is performed, the signals must be synchronized to eliminate the continually changing geometric delay of one antenna relative to another due to the Earth's rotation. For digitized signals large buffers of high-speed memory are used to delay the signals.

Even after crude digitization the processing rates required in the correlation step far exceed those feasible even with the fastest supercomputers. This problem can be exacerbated by a need to correlate for a number of different temporal lags, which is necessary if either the geometry of the interferometry is poorly known or spectral information is desired. For typical sample rates of 107 samples/sec for each of 10–30 antennas and up to 512 different temporal lags, the required multiplication rates can approach 1014/ sec. Special-purpose hardware is vital, and much effort has been expended in this direction. Most correlators use custom VLSI chips for the crucial multiplier-accumulator unit. For arrays of many antennas, the conventional correlator design of many multipliers and accumulators, one per lag per antenna pair, is extremely inefficient. An FFT chip to transform the signals to the frequency domain before multiplication is often preferable as it saves in digital logical and can also allow elimination of interfering signals.

Earth rotation, while helpful in sampling regions of the u−ν plane, complicates correlation. First, as a given pair of antennas traces out its ellipse in the u−ν plane, fine structure in the coherence function will be smeared out unless the integration time is short (∼10 sec). Second, the relative motion between the antennas introduces a differential Doppler shift in the received radiation, which must be cancelled. The measured coherence samples are usually averaged for as long as possible, within the limits posed by tolerable smearing of the coherence function or by uncertainties in the interferometry geometry.

The Earth's atmosphere can be a major source of error in the final coherence samples. At centimeter wavelengths the effect of the troposphere can be ignored for resolutions poorer than about 1 arcsec. For higher resolutions the averaged coherence samples must be calibrated by interleaving observations of an object of known strength and position nearby in the sky. In most cases even this procedure is not sufficient to allow high-quality imaging because the atmospheric variations will be partially decorrelated over even small antenna separations. The best remedy is then to “self-calibrate” on the source of interest itself (see Section III). At long wavelengths (>0.1 m) the ionosphere is the dominant cause of phase errors, but self-calibration can also remove these effects.

High-quality imaging requires good sampling of the coherence function. The largest separation of the antennas fixes the highest resolution possible, while the distribution of the samples over the u−ν plane determines the complexity of the structure than can be imaged. As the information collected scales as N2, it is desirable to have as many elements as possible. With a fixed budget one must include consideration of a number of factors including the number and diameter of antennas, cost and feasibility of the correlator, and ancillary computing costs. Practical constraints on the placement of antennas include the availability of land, the signal distribution system, and the mobility of antennas.

Table I summarizes salient details about radio-astronomical arrays currently in operation around the world.

TABLE I.

11.5.3 The Spectral Representation

Next we shall use spatial spectra to examine the effects that the weighting functions in Eq. (11.122) have on received power. Application of Eq. (10.7) to scalar fields allows Eq. (11.122) to be written as

(11.126)ℑ=2σr σ⊥2π3/2∫∫−∞ +∞∫Φn(K)[∫∫−∞+∞∫exp[j(K −2k0az0)·ρ]H(ρ)dVρ]dVK,

where

(11.127) H(ρ)=exp[−δz28 σr2−π2σ⊥2f4( δx2+δy2)]

is the weighting function in lag space. We have omitted the beamwidth term because, in the antenna's far field, it is significantly smaller than the Fresnel term. The bracketed term in Eq. (11.126) is a sampling function, similar to that of Eq. (11.81) but modified by the weighting function H(ρ). We therefore define a normalized spectral sampling function

(11.128)F(K)≡18π3∫∫ ∫exp[j(K−2k0az0 )·ρ]H(ρ)dVρ,

which has a peak value at K = 2k0az0. As the width of H(ρ) becomes broader, the spectral sampling function becomes narrower. The integrations along ρ can be performed to give

(11.129)F(K)= 0.44D2σrln2π7/2exp[−2σr2(Kz−2k 0)2−D2(Kx2+Ky2)ln23.24π2],

in which we have substituted Eq. (11.123) for σ⊥. Thus, the larger the antenna diameterD, the narrower is the sampling function. A surprising result is that the sampling function shape is independent of r0. For a given antenna diameter, the spectrum Φn(K) of irregularities is weighted equally for all resolution volumes in space. Condition (11.119c) however, must be satisfied.

To provide physical insight, we now consider a spectral description for the angular dependence of scatter from anisotropic irregularities. From Eqs. (11.126), (11.128), and (11.116), the backscattered power becomes

(11.130)〈Pt〉= 22(0.45)2Ptg2σr π9/2r02D2ln2∫∫∫Φn(K)F(K )dVK.

Equation (11.130) shows that the echo power is proportional to the integral of the product of the spectral intensity Φn(K) of the refractive index irregularities and the normalized sampling function F(K). Now, if the horizontal correlation lengths are large compared to the vertical ones, Φn(K) will be sharply peaked in the Kx, Ky directions and less peaked along the Kz axis. If the irregularities can be roughly described as oblate spheroids, then the correlation R(ρ) has a similar form, but Φn(K) is prolate spheroidal in shape (Fig. 11.13a).

In deriving F(K) we assumed the z axis to be along the beam axis but made no assumption as to the direction of z in space relative to the earth. Furthermore, the shape of the sampling function depends only on the range resolution and antenna diameter and is thus independent of the location of the resolution volume V6 in real space. Therefore, in K space F(K) is invariant under rotation of the K coordinate axes, and the sampling function (11.129) can be formulated more generally as

(11.131)F(K)=0.44D2σrln2 π7/2exp[−2σr2(K||−2k0)2−D2K⊥2ln23.24π2],

where K|| is the wave-number magnitude along k0, and K⊥ is the perpendicular distance of K from k0. Thus, if z is along the vertical and the beam axis direction (i.e., k0/k0) is rotated by π/2 – θ from it, then F(K) must be rotated by π/2 – θ from the Kz axis. Equation (11.131) reveals that whenever the range resolution r6 = 3.33σr > 0.34D, a common situation, F(K) will be narrower along K|| than along K⊥. Figure 11.13b shows the contours of F(K).

Obviously, we obtain maximum for F(K) centered where Φn(K) is a maximum. But, we do not have the freedom to choose arbitrarily the location of F(K), because the radar wavelength centers the F(K) peak at a wave-number magnitude 2k0. For the conditions shown in Fig. 11.13, maximum occurs when θ → π/2. The angular dependence of on θ is strongly influenced by the sharpness of Φn(K) and F(K). If Φn(K) is highly anisotropic (as in Fig. 11.13a) then will be linearly dependent on the range resolution but proportional to the square of the antenna diameter D, because g2 is proportional to D4 and the integral in Eq. (11.130) is independent of D.

To measure Φn(K) accurately with radar requires F(K) to be sharp compared to the shape of Φn(K). Thus, for anisotropic irregularities we need D > ρh and σr > ρv, where ρv is the vertical correlation length. For spectral samples along the horizontal direction, we must have D > ρv and σr > ρh. That the antenna diameter instead of the linear beamwidth (r0θ1) enters into Eq. (11.131) illustrates the importance of the second-order expansion (11.119a) and the Fresnel term. Only when this term is negligible does the sampling function narrow as the linear beamwidth widens. Furthermore, the Fresnel term results because the irregularities have their axis of symmetry along Cartesian coordinates, whereas the phase fronts of the radar are spherical. If the irregularities had shapes that were concave downward with radius of curvature equal to r0 (this occurs in an ideal RASS, Section 11.4.2), then there would be no Fresnel term, and the width of F(K) would be inversely proportional to the linear beamwidth as well as to the range resolution.

1.2.3.2 Signal parameters

From the above discussions, we come to know about the metrics which quantitatively define the quality of service of a satellite system. Here, we shall discuss about the signal parameters which directly affect these terms.

The BER, Availability and Continuity terms are determined by the ability with which the receiver can acquire the signal and can correctly identify and track the current states of the signal parameters, viz. amplitude, phase, frequency, etc., which bears the information in form of bits. Only on correct identification of the bits that the receiver can faithfully recreate the information embedded in it. It is obvious that the propagation factors will distort the signal at the receiver from its condition at the transmitter. The signal is designed and transmitted in such a way that the performance targets are still achieved with the designated receivers for the expected conditions of propagation. So, under nominal condition, there is no scope for any difficulty to happen in carrying out the communication processes with expected performance. But, the pragmatic scenario may be different and the signal may actually be received in much more distorted form than what is otherwise expected with consequent divergence from its predicted performance. In fact, there are candidate factors which distort the signal character and thus deteriorate the signal quality as it passes from the transmitter to the receiver. The signal power may be excessively reduced, some meaningless random voltages, called Noise may get added with the actual signal levels or similar degrading effects may happen. The effects may be so severe that the information in the signal may get totally unintelligible at the receiver. However, the more rigorously the propagation factors are considered in the design, the more rugged the signal is to the propagation impairments. Here, we are now going to learn about the signal parameters which have explicit bearing upon the performance.

The information-bearing units in the signal are called bits. Binary bits can assume values ‘0’ or ‘1’. Bits, in a carrier signal, are represented by predetermined state of the signal parameter. For example, in quadrature phase shift keying (QPSK) type of modulation, the four possible codes, corresponding to two bits of information, which are (0,0), (0,1), (1,0) and (1,1), can be represented by four different phase states of the signal. The states are typically separated by equal phase differences. So, with reference to an arbitrary zero phase of the unmodulated carrier signal, two of the codes may be represented by the phase + π/4 and + 3π/4, while the other two codes may be represented by phase − π/4 and − 3π/4 of the signal.

At the receiver, these bits are identified from the signal states. Correct identification of the bits is determined by the ability of the receiver to distinguish between these states of the signal representing different bits. This distinction is made by comparing the signal state with a predetermined reference state. For a given separation of the states, the receiver is able to correctly distinguish them and decide the correct bit when more energy is carried by the actual signal within the duration of a bit. Moreover, increased contribution of the noise in the signal increases the likelihood of incorrect identification of the information bit. Therefore, the overall probability of a bit to be identified correctly depends upon how much the different states are separated from each other, and also upon how much energy is carried by the signal within the interval of a bit (Eb). It is also inversely proportional to the variance of the noise. The total added noise is a linear function of the noise power spectral density, N0. Hence, the ratio Eb/N0, is one of the most important parameters in satellite communication system which determines the BEP or BER. We can write this term as

(1.1)EbN0= S×TbN/B=SN× BRb

where, S and N are the power of the signal and the noise, respectively, and Tb (=1/Rb) is the bit duration. Rb is the bit rate and B is the bandwidth of the signal. The ratio S/N is called the signal to noise power ratio or simply signal to noise ratio (SNR). Here, we have considered the fact that the noise power spectral density N0 is constant over the bandwidth, B. Rearranging, Eq. 1.1 can also be written as

(1.2)SN=EbN0×RbB

Therefore, S/N and Eb/N0 are related by the ratio Rb/B, called as the bandwidth efficiency. It represents the bit rate that can be achieved given a definite bandwidth of the signal and depends upon the type of modulation. So, for the same requirement of Eb/N0, the S/N requirement will be different depending upon the modulation. Further, as the signal to noise ratio is derived from the RF signal, with C and N as the respective powers in the carrier and noise, we can write,

(1.3)EbN0=C/NRb/B=CN0×1Rb

If, there are more than one information channel in the same carrier, the total carrier power is accordingly apportioned. This division of power is accredited to individual channel at the time of transmission. Eb/N0 of each channel is then estimated with its respective carrier power and can be written as

(1.4)EbN0j=CN0j×1Rbj

As BER improves with increasing Eb/N0, the improvement is observed with both increased (C/N0) and with increased Tb, the latter being equivalent to reduction in Rb. While BER is a measured number, the relationship of Eb/N0 exists with the BEP. The exact mathematical expression relating the BEP with the Eb/N0 can be obtained from any good book on digital communication (Proakis and Salehi, 2008; Chakrabarty and Datta, 2007; Mutugi, 2012). Table 1.2 provides the final expression for the BEP in terms of the Eb/N0 (Maral and Bousquet, 2006; Mutugi, 2012).

Table 1.2. Expressions for BEP for different modulations

It is worth reiterating that, in order to retrieve the data, the receiver needs to acquire the signal first. Acquisition is the detection of the approximate carrier states and adjustments of the local reference signal of the receiver for demodulation. Once the signal is acquired, it is to be tracked throughout the period of its operation. Tracking is nothing but following the changes in the incoming signal states and accordingly adjusting the local reference signal so that the demodulation process is correctly continued. Both acquisition and tracking performance, which directly affect the availability and continuity of the services, improve with better values for C/N0.

It is evident, thus, that higher the C/N0 of the received carrier, better the probability of the receiver to acquire and track a signal and identify the information-bearing bits in it correctly. Hence, the quality of the satellite services, which depends upon the correctness of the received signal, is determined by the received carrier power to the noise power density.

The term C/N0 of the received signal is thus found to be the most important parameter influencing the performance of a communication system. Now, let us understand how this factor is modified during propagation. We shall deal with modest details here and will recall some of the fundamental concepts discussed here again in Chapter 9. The power of a signal received by a receiver depends upon the transmitted power PT. This power when transmitted by a directive antenna, having directive gain G, the total power is concentrated along the direction of the gain of the antenna. This enhanced power in the direction of the antenna gain is called the effective isotropic radiated power (EIRP) of the signal. Therefore,

(1.5)EIRP=PT×G

The enhanced power is then transmitted towards the targeted direction. On propagating through the space, the power of the signal gets weaker as it spreads over larger and larger area. At a distance R, the power passing per unit cross sectional area is the power flux density, given by

(1.6)ΦR=EIRP4πR2

Every receiving antenna has an effective receiving area determined by its gain such that the power incident over this area can be received by it. If GR is the gain of the receiving antenna and Ae is its effective area, then the relationship is given by

(1.7)GR =4πλ2AeAe=Gλ 24π

Here λ is the wavelength of the signal. Therefore, the power received by the receiving antenna with gain GR is

(1.8)PR=ΦRAe=EIRP4πR2×GRλ24π=EIRP4π R/λ2×GR

Calling (4πR/λ)2 as free space path loss (LFS), the received power after being enhanced by the receiving antenna gain becomes,

(1.9a)PR=EIRPLFS×GR

The absolute power level variation from the EIRP to the received power is very large, and hence, is more conveniently represented in logarithmic scales. So, in logarithmic (decibel) scale,

(1.9b)PR=EIRP–LF S+GRindB

In addition to the free space loss occurring due to the obvious spreading of the power with distance, there is also some associated loss due to the passage of the signal through the atmosphere. It is called the Atmospheric propagation loss, LA. Considering the latter, the received power becomes

(1.9c)PR=EIRP–LFS–LA+GRindB

To this received power, some unwanted and uncorrelated power from random noise will get added. These are generated due to the emissions of incoherent electromagnetic wave by the atmospheric elements at finite temperature and which are picked up by the antenna. Further added to this will be the noise generated by the random motion of the electronic charges in the antenna and the receiver hardware. This noise component gets adhered with the actual signal and degrades the signal quality and thereby reducing the ability of the receiver to correctly identify the signal state.

The occurrence of noise in time domain is fully random and its probability density follows Gaussian distribution with zero-mean. The power spectrum of noise is white, i.e. noise has almost the same spectral power over the frequency band width of interest. A black body at a finite temperature will also emit noise over all frequency bands. The spectrum of the received power from such emitted noise is approximately uniform in the radio frequency range and is only dependent upon the temperature of the body. Therefore, the noise added to the signal may be assumed to be transmitted by a black body at a finite temperature, T. This is called the equivalent noise temperature. Since the power contribution over the entire spectral range is equal, the noise is characterized and better represented by noise spectral density N0. This is the total noise power over a unit bandwidth. As this is proportional to the noise temperature, we can write N0 = kT. The proportionality constant ‘k’ is called the Boltzmann constant and its value is 1.38 × 10− 23 J/K or − 228.6 dB. Thus, the terms N0 and T may be used interchangeably. Therefore, the ratio of the carrier power to the receiver noise density may be expressed as

(1.10)C /N0=EIRP–LFS–LA+GR−kTindB=EIRP–LFS–LA+GR/T−kindB

All other parameters being factors of transmitter and propagation medium, the performance of the receiver terminal is measured by the term GR/T, the ratio of the receive antenna gain and the system noise temperature. It is called the figure of merit of the receiver.

Box 1.1

MATLAB Exercise

The MATLAB Link.m was run to generate the following variations of the received C/N0 for different frequencies of the satellite signal. The given input and the program output for this particular run are shown below. The difference in C/N0 is essentially due to the difference in path loss, all other factors remaining the same. A considerable difference is observed between Ku and Ka band signals.

Signal parameters:

Input the transmission frequency (GHz): (12:2:24)

Transmitting ground station parameters:

Input the transmission power (W): 6

Input the ground station antenna diameter (m): 2.4

Input the ground station antenna efficiency: 0.55

Input the ground station latitude (deg; E: + ve, W: − ve): 22

Input the ground station longitude (deg; N: + ve, W: − ve): 88

Receiving satellite parameters:

Input the satellite antenna diameter (m): 1.2

Input the satellite antenna efficiency: 0.6

Input the satellite longitude (deg; E: + ve, W: − ve): 83

Input the noise temperature at receiver (K): 450

The estimated parameters are:

The ground station antenna gain for the range of input frequencies is respectively (dBi)

The ground station EIRP for the range of input frequencies is respectively (dBw)

The satellite-ground station distance is 36,568.7602 km

The power flux density for the range of input frequencies is respectively (dBw/m2)

The satellite received power for the range of input frequencies is respectively (dBw)

The noise at the satellite is − 202.0679 dBw

The carrier to noise ratio for the range of input frequencies is respectively (dBHz)

Input the minimum Eb/N0 required for your modulation (dBHz): 14

The maximum bit rate possible for the range of input frequencies is respectively (Mbps)

Run the program for different positions of the satellite, different G/T and for different EIRP and compare the graphs. The atmospheric loss is considered to be absent here. The program takes only frequency values in an array.

For satellite-based broadcasting, navigation, remote sensing and certain other applications, communication of data takes place in one direction between the satellite and the ground terminal. Therefore, either the uplink or the downlink is in consideration. There, the link performance is obtained from the one way estimate of C/N0 as given in the equation mentioned previously. Nevertheless, in a communication system, where the link is established between two ground terminals through a transparent satellite, the total link between these two ground terminals will consist of an uplink and a downlink connection. The satellite will receive the signal from the transmitting ground station and convert it to a different frequency before transmitting it again towards the receiving ground station after amplification. For such a communication link, the total C/N0 of the signal over the complete path at the receiving ground terminal, represented by (C/N0)T, is given by

(1.11)CN0T=CN0U− 1+CN0D−1−1

where (C/N0)U is the carrier to noise density ratio for the uplink and (C/N0)D is the carrier to noise density ratio for the downlink part of the total connection, considered individually.

From the expression in Eq. (1.1), it is clear that (C/N0)T is even lower than the lowest among the values of (C/N0)U and (C/N0)D. Therefore, if only one of these values degrades to fall below the threshold needed to ensure a certain performance level, the effective (C/N0)T cannot be raised above this threshold just by increasing the C/N0 of the other side. Therefore, to keep the (C/N0)T above a particular required value, both the uplink and the downlink components of it are individually required to be maintained above it. However, if one of these components is kept at a sufficiently larger values than the other, then the (C/N0)T becomes approximately equal to the other component. The following Focus 1.2 describes how to calculate the C/N0 of a link considering all the pertinent factors. This will help in understanding where the propagation effects are involved and by how much. Similar problems will be dealt with in Chapter 9.

Focus 1.2 Satellite Link C/N0 Calculation

We have already seen how the signal transmitted from the ground is received by the satellite and is retransmitted back to the ground after amplification. Given the ground station and the satellite trans-receive parameters and their distances, let us find the C/N0 values at the destination receiver.

The system design parameters are:

Among the ground station parameters, assumed to be identical for transmitting and receiving:

The satellite design parameters are

The wavelength for the frequency used is

λ=3×108/30×109=10−2m

The transmitting antenna gain is given by

GT= 4π/λ2⁎πD2/4⁎η=9.4748 ×105=59.76dBj

Therefore, the EIRP is given by,

EIRP=P×GT=9.4748×106W= 69.76dBW

The power flux density at the distance of the satellite is given by

φ=EIRP/ 4πd2=4.7124×10−10W/m2

The effective area of the satellite receiving antenna is

Aes=πDs2/4×ηs=2.1991m2

The carrier power received by the satellite is

Ps=φ×Aes =1.0363×10−9W=−89.84dBW

Again, using the constant value of k, the noise density at the satellite receiver is

Ns=k×Ts= 6.2117×10−21=−202.07dB

Therefore, the (C/N0)U becomes

C/N0U=Ps/Ns= 1.668×1011=112.22dBHz

In this variable, the propagation factor is reflected in the received power flux density.

Now, the signal is amplified at the satellite transponder after the carrier frequency is appropriately adjusted. But, once the noise gets added to the signal, the signal to noise ratio cannot be improved just by amplifying the combined signal. This is because the noise also gets simultaneously amplified as a result, keeping the ratio unaltered. Therefore, if the carrier is amplified by factor A at the satellite transmitter to become C × A, the noise power density there due to the already added noise during the uplink is also N0 × A. Taking satellite transmitting antenna same as the receiving antenna, with gain Gs,

Gs=4π/λ2×Aes=2.763×105= 54.41dBi

EIRPs=6×2.763×105 =16.57×105W=62.19dB

The nominal C/N0 at the ground receiver with Tg = 250 K is

C/N0g= 1.8×1011=112.55dBHz

Now, the received carrier power at the destination ground station receiver

Cd=C×A×Gs/d2×Aeg

where d is the radial distance from the satellite and Aeg is the effective antenna area of the ground station. Similarly, the noise power received here due to the noise already added with the signal during uplink is

N0u=N0×A×Gs/d2×Aeg

If the noise temperature at the destination ground station is Tg = 250 K, the C/N0 for the downlink is

C/N0d=C×A×Gs/d2×Aeg/kTg

Now, during the RIP from the satellite to the ground, the signal also acquired some noise power. The power density of this noise being N0d, we get the total noise power density at the ground receiver as

NT=N0u×A×Gs/d2×Aeg+ N0d

Therefore, the C/N0 after the round trip of the signal is

C/N0T=C×A×Gs/d2×Aeg/N0u×A×Gs /d2×Aeg+N0dOr ,C/N0T−1=N0u×A×Gs/d2×Ae g+N0d/C×A×Gs/d2×Aeg=N0u/ C+d2×Aeg×N0d/ Cu×A×Gs=C/N0u −1+N0d/Cd=C /N0u−1+C/N0d− 1

Using the obtained values, we get

C/N0T=8.659 ×1010=109.375dB

Apart from the information sent through the coded message in a signal, some applications also need the signal propagation parameters like the range travelled by the signal, etc. to be derived in situ by the receiver from the received state of the signal itself. In such satellite applications like altimetry, navigation or radar systems, this derivation is typically based upon the time delay between the transmission and reception of the signals and hence in turn depends upon the nominal velocity of the signal. Therefore, any additional delay added by the medium during the propagation causes error to these applications. In such applications, the additional signal delay caused by the propagating medium is of prime importance. This additional delay is caused by the deviation in the nominal refractive index of the medium which in turn is again a function of its constituents.

So, to round the things up, we have found that the performances of the satellite-based system for different applications are governed by the signal parameters like the received signal power or more precisely by the received C/N0. The time delay of propagation, etc. also influences the performance in certain applications. We have also seen that the received C/N0 is a function of the atmospheric path loss, while the delay is dependent upon the refractive index (RI) of the medium and hence in turn to the atmospheric constituents. In our later chapters, we shall see in details how each of the propagating factors affects the C/N0 term. We shall also see, how much is the effect of the degradation of the C/N0 on the signal acquisition and data retrieval processes and how we can improve or compensate the deterioration of this term. The reasons for the signal delays will also be explored and how to compensate them will be discussed. However, before going to that, the following section will give us an overview of the atmosphere and its role in causing propagation impairments.

II.B.1 Antennas

Satellite antennas concentrate the satellite's transmitting power into a designated geographical region on earth and avoid interference from undesired signals transmitted from outside of the service area. The antenna radiation characteristic is usually specified by its half-power beamwidth (HPBW). For an aperture antenna, HPBW is approximately given by

(21)θHPBW=65 ∘λD

where D is the antenna diameter and λ is the wavelength. The relationship among the HPBW, the earth coverage diameter, and the antenna diameter at different frequencies is shown in Table VI. Global beam coverage from geostationary satellites needs a field of view of 17.4°. Since reflector antennas are not efficient when the diameter is less than 8λ, horn antennas or wire antennas are used for a wide-beam antenna whose beam angle is larger than about 8°. On the other hand, the larger reflector antennas are needed to produce spot-beam coverage in the lower frequencies.

Four main types of antennas are used on present communication satellites: wire antennas, horn antennas, reflector antennas, and array antennas. Wire antennas were used on the early operational satellites such as INTELSAT I and II with an antenna gain of about 4 dBi for receive and about 9 dBi for transmit. Now they are mostly used as the tracking, telemetry, and command (TT&C) antennas because of their wide beamwidth.

Horn antennas are one of the simplest directional antennas. Depending on their shape, there are several types of horn antennas: pyramidal, sectoral, conical, dual-mode, corrugated, multiflare, etc. (Fig. 15). The pyramidal horn and conical horn are simple and easy to fabricate, but they have the drawback that the radiation pattern is not circularly symmetrical and so antenna efficiency is low. They are also used as array elements of array antennas because of their small sizes. The dual-mode horn can produce a circularly symmetric pattern and lower sidelobes by adding TE11 and TM11 modes, but the disadvantage is narrow bandwidth. The corrugated horn generates a circularly symmetric pattern, low sidelobes, and low cross-polarization by using the hybrid electric HE11 mode. These dual-mode, corrugated, and multiflare horns are also used as feeds for reflector antennas.

The reflector antenna is most frequently used in communications satellites because of its simple structure, light weight, and high gain (Fig. 16). It has one or more reflective surfaces, which are paraboloid, hyperboloid, spheroid, etc. A parabolic antenna consists of a single reflector shaped as an axially symmetric paraboloid of revolution and the feed situated at the reflector's focus. The feed may block some of arriving waves and cause antenna gain to drop and sidelobes to increase. The reflector of an offset-parabolic antenna, therefore, has offset to avoid the aperture blockage. A Cassegrain antenna uses a paraboloid of revolution as a main reflector and a hyperboloid of revolution as a subreflector. One focal point of the subreflector coincides with the focal point of the main reflector, while the other focal point of the subreflector coincides with the feed. Multibeam using reflector antennas can be achieved by multiple feeds situated near the focal position. An increase in the number of feeds, however, causes degradation of the radiation performance because offset of multiple feeds from the focal position increases. This type of multibeam antennas is used in offset parabolic and offset Cassegrain antennas.

An array antenna can generate steerable beams and particular radiating patterns with high directivity and low sidelobes by using a large number of radiating elements. The design of an array involves the selection of radiating elements and array geometry and the determination of the element excitations required for achieving a particular performance, which is not possible with a single radiating element. As the radiation elements in a satellite phased array antenna, horn, dipole, helix, and microstrip patch antennas are mostly used. Figure 17 shows an S-band phased array antenna, which was installed in the ETS-VI satellite to establish intersatellite links between the geostationary satellite and low-earth-orbiting satellites. The array consists of 19 radiating elements, each of which is equipped with one phase shifter for transmit beams and two phase shifters for receive beams so that each of these beams can be electrically scanned independently. This antenna is 1.8 × 1.8 m in size. Figure 18 shows a Ka-band active phased array antenna (APAA) of the Japanese Gigabit Satellite. Two antennas with an aperture diameter of 2.2 m for transmit and 1.5 m for receive are installed on the same surface of the satellite's earth panel without deployable structure. The APAA consists of 38 subarray units, each of which consists of 64 horn-antenna elements with a mutual spacing of 2.2λ. Figure 18 shows this subarray unit. Four beams with a scanning angle of ±8° are available.

To allow larger aperture and larger gain of satellite antennas, a deployable antenna, which uses meshed or solid reflectors folded for launch and deployed on orbit, is needed. There are two types of deployable antennas: the wrap-rib type and the umbrella type. The wrap-rib antenna consists of flat ribs that can spread out into radial directions, a center hub, and a reflective film (mesh) that extends across the ribs. The umbrella-type deployable antenna also consists of multiple ribs and a center hub, but it can achieve expansion by having the ribs open up from their base like an umbrella. Figure 19 shows multiple umbrella-type reflectors installed in the ETS-VIII satellite. The unfolded reflector size is 19.2 m × 16.7 m. The reflector is assembled with 14 umbrella-type modules, each of which is deployed by using stepping motors. To achieve a multibeam phased-array antenna, the feed consists of 31 elements. The feed array is located about 1 m away from the focal point. Two separate reflectors for transmit and receive in S band are installed in the satellite to avoid signal coupling and interference by passive intermodulation.

IX.A Types of Lunar Observatories

Of the many types of observatories suggested for the lunar surface, four are mentioned here to illustrate the scope of the engineering challenges.

The concept of a Moon–Earth radio interferometer (MERI) suggested by Burns would include radio antennae on the Moon. An advantage would be a resolution (at the 6-cm wavelength) 30 times better than the very long baseline array (VLBA). The MERI could begin with a 10- to 15-m diameter antenna on the Moon and progress to a larger antenna and then an array of antennae on the lunar surface. Large parabolic dish antennae on the Moon will be designed for robotic erection and ease of maintenance. They must be able to maintain shape in the wide temperature changes of the lunar surface. The structural aspects will require careful design. On the Moon, lack of wind loads and less gravity loading will permit use of less massive structures than their very large array (VLA) counterparts near Socorro, New Mexico.

Douglas and Smith of the University of Texas have proposed a simple very low frequency (VLF) radio telescope for the lunar surface to investigate the radio sky beyond the 10-m wavelength. This portion of the radio sky is relatively unknown because of terrestrial ionospheric absorption. Burns and co-workers suggest a 20-km diameter array of dipoles involving short wires laid on the lunar surface. Each wire would be equipped with an amplifier and digitizer communicating with a common computer. It would accomplish high-resolution observations in the 10- to 100-m wavelength range and lower resolution observations up to the 1000-m wavelength. Such an array would be designed for ease of placement. Reliance would be on the ability to transverse with relative ease a large area on the lunar surface and to site, design, construct and shield a central computer facility. A potential location would be in the Crater Tsiolkovsky on the far side of the Moon away from terrestrial radio interference according to Taylor.

Optical interferometry from the lunar surface has been suggested by Burke of MIT, who noted that if the technical challenges can be met, a resolution of nearly 1 μ arcsec could be achieved with a wye-shaped array of 27 one-meter optical telescopes (Fig. 6a). Atmospheric effects preclude the operation of such a system on the Earth. Operating such a system in Earth orbit would be difficult because of stringent requirements on controlling element position and orientation to about 100 Å in 20 km. The usefulness of the optical interferometer rests on the extension of the radio astronomy concept of aperture synthesis. Each arm of the wye would be 6-km long, and the received light from the nine telescopes on each arm would be transmitted to a central correlation station through a set of variable time delays.

A very large Arecibo-type (Fig. 6b) telescope has also been suggested for the lunar surface as well as sets of instruments for X-ray and γ-ray astronomy, and infrared astronomy. Table XII lists some design considerations for four observatory options. For each option, there are common design considerations, such as making use of lunar materials (e.g., for shielding in the near term, and for manufacturing composite structural materials in the far term), minimizing mass to be transported from Earth, packaging for transport, and reducing erection complexity. Each of these four suggested observatory installations poses a set of problems to the designer. Design must take into account many factors including the unique nature of the lunar environment.

TABLE XII. Examples of Design Considerations for Four Observatory Options

6.3.2 Architecture very small aperture terminal networks

VSAT networks have various characteristics, such as operating in one or more frequency bands in the exclusive area of the following bands allocated to FSS (Fixed Service Satellite): 14.00–14.25 GHz (Earth to space), 12.50–12.75 GHz (space to Earth), or in the shared areas of the following bands allocated FSS 14.25–14.50 GHz (Earth to space) and 10.70–11.70 GHz (space to Earth). All these networks operate with geostationary satellites, and terminals are provided for unattended operation, with an antenna diameter upto 3.8 m.

The European Telecommunication Standards Institute provides specifications for standardizing the characteristics of VSAT networks operating as part of a satellite communications network (in star, mesh, or point-to-point configuration).

The network is coordinated by a ground station called the hub. The architecture is generally a star type and offers great flexibility through single-hop as well as double-hop connections. In the case of a single-hop connection, the data is transferred between the hub and a VSAT terminal. The double-hop connection allows data transfer between two VSAT terminals through the hub. In the case analyzed in this chapter, the satellite connection is of the double-hop type; the communication path between the SCADA location and the Main/Secondary Dispatcher is SCADA location -> satellite -> central hub communication by the VSAT satellite of the operator -> satellite—> Main Dispatcher.

Direct connections between VSAT terminals are possible if the power received by the satellite is sufficient and at the same time if the satellite has sufficient power to relay the signal. The system is economical because VSAT stations have antennas of small diameters (below 2.5 m) and low emission power. Therefore direct connections between stations are impossible. The hub is equipped with a large antenna, capable of receiving the low power signals of a VSAT, relayed by the satellite. It transmits or retransmits high-power data so that they can be received using the small antennas of the VSAT terminals. The hub is used for routing traffic and ensures network switching.

In conclusion, the hub performs processing and switching functions that are not available on a satellite, which is generally a classic repeater.

VSAT networks operating in 4–6 GHz bands face serious problems regarding the interference with neighboring satellites and terrestrial microwave networks that use the same frequency range. This has required the use of scattered spectrum transmission techniques, with the satellite transmission capacity being used less efficiently. Interferences are reduced in the 12–14 GHz bands; here the connections are affected by atmospheric phenomena, in particular rain (depolarization, attenuation).

In the double-hop mode, one VSAT terminal accesses another terminal through the hub. All signals are received by the hub, which acts as a processor, decoding, demultiplexing, regenerating, multiplexing, encoding, and transmitting all data transmitted in the network.

Typical modulations used in this type of network are binary phase-shift keying and quadrature phase-shift keying. The multiple access modes are the ones known as time division multiple access (TDMA), frequency division multiple access (FDMA), code division multiple access (CDMA), and demand assignment multiple access (DAMA), depending on the particularities of the network.

VSAT is a satellite system for home or business users, so by connecting the VSAT terminals to a terrestrial hub, a network can be created, at low costs, to provide communications only through this hub (STAR network configuration, Fig. 6.3).

The satellite network subsystem is made out of a main antenna connected to the hub, a satellite, and numerous smaller (spaced between themselves) terminal antennas.

At the customer’s location, the VSAT system includes a receiver (satellite router), low noise converter, frequency converter (block upconverter), and a parabolic antenna [17].

The terrestrial network subsystem consists of the links that lead from the user’s central block to the hub. This network made of specialized links has all the characteristics of a classic network and is configured accordingly.

The hub converts the protocols and carries the necessary information flow to the recipient on the satellite channel, being the main element of the communications system, containing elements through which input/output (I/O) connections are created to and from the terminals (Fig. 6.4).

A hub may be passive (when performing the same function as a repeater [18]), smart (when it has built-in software to configure each port in the hub and to monitor the traffic passing through it), or switching (when reading the destination address of the packet and directs it only to the port indicated by the address).

The terminal station comprises three elements: the antenna, the external radio unit, and the internal radio unit. In the case of the VSAT network, the antenna has an average diameter of upto 2 m (a diameter smaller than 1 m will reduce the data flow below 9600 bauds) [19].

The VSAT antennas are equipped with an external radio unit (located at the top of the antenna) that allows data transmission and reception. The internal radio unit can be a modulator/demodulator of the signal taken by the antenna and introduced into the computer system for information processing.

The transceiver sends and receives signals from a transponder situated on a satellite. By transponder (Transmitter-Responder) we understand the unit receiving–broadcasting from telecommunication equipment (this includes other components, including power, control, and cooling units); transponders are on satellites but also in some radio-relaying systems.

The satellite network is traversed by two data streams (Fig. 6.5A): a data stream called outbound, whose direction is from the hub to the antenna, and a data stream called inbound, whose direction is from the antenna to the hub. Each of these data streams is at its turn composed of two different secondary streams (Fig. 6.5B): an upward data stream situated on the lifting channel (uplink), which links the hub to the satellite, and a downstream data stream on the descending channel (downlink), which connects the antennas to the hub.

Different frequencies are used for lifting and lowering to prevent transponders from entering the oscillation. Therefore the VSAT network must be seen as a juxtaposition of two totally different unidirectional subnets: the first is an “outbound” network, going from the hub to the VSAT station, and the second is an “inbound” network from the stations to the hub.

TDMA is a technique used in the topology of the star satellite network, where all remote stations are dynamically connected (in time division) via satellite to a central hub [20]. Two or more terminal locations can communicate with each other through the central hub, with double delay round trip (Fig. 6.6A). This network topology is extremely flexible and cost effective, being the ideal choice for the type of networks that, like SCADA networks, must transmit real-time information. VSAT terminals that are farther from the satellite, such as site 1 (Fig. 6.6B), must transmit earlier than site 2, which is closer to the satellite.

A transmission operation is carried out in three stages.

The first stage is the preparation phase, which consists of addressing all the stations to the list for transmitting and establishing a dialog with each of the places addressed and spaced between them.

Once this phase is completed, the actual transmission phase can be carried out, in which the communication is carried out without any control over the information flow. In principle, the transmitter sends the data in block form, together with a control block containing the error detector code and a sequence number. Each receiver will treat the received blocks by checking the error detector code and storing the valid blocks. Because in the case of a single transmission, there will be a probability that a large number of stations will not receive the blocks correctly, this operation will be repeated several times. At each new retransmission, the receiving stations focus only on the blocks that lack information.

The third phase consists of querying all the stations, one by one.

In the case analyzed in this chapter, in all SCADA locations, the VSAT satellite communication system consists of router type indoor unit, low noise block downconverter (LNB) type outdoor unit, parabolic antenna (in most locations, as well as at the Main and Secondary Dispatchers), and RF cable and connectors [21,22].

Depending on the required bandwidth, VSAT systems are a bit bulky (antennas that are 1–2 m in diameter) and easy to install. VSAT terminals with very small antennas (diameter less than 1 m) are intended for small volume traffic.

Radioastronomy

One of the most fruitful application of laboratory microwave spectroscopy over the last twenty years is the analysis of the molecular content of interstellar clouds. These clouds contain gas (99% in mass) which has been mostly studied by radioastronomy, and dust, whose content has been analysed mostly by IR astronomy. The clouds rich in molecular content are dense or dark clouds (they present a large visual extinction), with a gas density of 103–106 molecules cm−3, and temperatures of T < 50 K. At these low temperatures only the low-lying quantum states of molecules can be thermally (or collisionally) excited, i.e. rotational levels. Spontaneous emission from these excited states occurs at microwave wavelengths. In some warm regions of dense clouds (star formation cores) the absorption of IR radiation produces rotational emission in excited vibrational states. Other rich chemical sources are the molecular clouds surrounding evolved old stars, such as IRC+10216, and called circumstellar clouds.

In the 1980s and 1990s a lot of radiotelescopes were built, with large antennas (diameter = 10–30 m) and sensitive receivers in the millimetre and submillimetre range. More than 100 different molecular species were found in the interstellar medium (see Table 1) and, for some of them, various isotopic species were also detected. The identification of interstellar species is not easy because of the high density of lines in the spectra of some interstellar clouds. A millimetre wave spectrum of the Orion nebula is shown in Figure 11. This is owing to the richness of the chemistry in these clouds and also to the improved sensitivity of the latest generation of radiotelescopes. The characterization of the molecules present in these dense cloud requires a knowledge of the laboratory spectra. In some cases (C3H2, HC9N, etc.) the identification was first made in the interstellar medium, before laboratory evidence. Nevertheless in the case of HC11N, the highest membrane of the cyanopolyine series, interstellar detection was claimed at the beginning of the 1980s. This molecule was recently produced in the laboratory and its rotational spectrum does not fit the interstellar line. A search for HC11N with the new experimental data was at first unsuccessful but, finally, a deeper search confirmed the presence of HC11N in the interstellar medium. A lot of laboratory studies have been devoted to this family of molecules: the rotational spectrum of HC17N has been observed, and numerous hydrocarbons of the type CnH m, with n > m, have been produced in discharges and their spectra analysed.

Table 1. Interstellar molecules

The detection of isotopomers in interstellar medium is a source of information on the elemental isotopic ratio. Molecules containing the following atoms have been detected: D, 13C, 15N, 17O, 18O, 33S, 34S and 36S. The deuterated species are of particular interest because their abundances bring useful information on the chemical processes which take place in the peculiar conditions of the interstellar medium (isotopic fractionation).

Molecular hydrogen is the dominant molecule; the second most abundant molecule, CO, is four orders of magnitude less abundant. But H2 has no strong transitions in the microwave regions, CO is mainly used to map interstellar clouds in our galaxy and others, and also in quasars. The observation of several lines of the same species gives information on the physical conditions in the interstellar cloud: temperature, molecular density. In the case of OH radical, the splitting of the observed microwave lines by the local magnetic field (Zeeman effect) is a way to evaluate its order of magnitude. Several molecular ions have been studied in the laboratory (H2D+, H3O+, CH2D+, etc.) because of their importance in interstellar chemistry, which consists mostly in gas phase ion–molecule reactions. But in many cases their reactivity prevents their interstellar detection. Radioastronomy has also been applied to the analysis of planetary atmospheres, together with infrared observations. Both CO and H2O were detected in Mars and Venus, SO2 in Io (a satellite of Jupiter), CO and HCN in Neptune. In Titan, a satellite of Saturn, HCN, HC3N and CH3CN were detected, indicating a complex photochemistry. More detailed mappings were undertaken more recently with interferometers working in the millimetre-wave region.

Millimetre astronomy has also been found to be a powerful tool for the physicochemistry of comets. This was fully demonstrated by the observations of two exceptional comets: Hyakutake (1996) and Hale–Bopp (1996–1997). The newly detected molecules in these two comets are: CS, NH3, HNC, HDO, CH3CN, OCS, HNCO, HC3N, SO, SO2, HCCS, HCOOH, NH2CHO, CN, CO+, HCO+ H3O+. This number is considerably bigger than the total number of molecules previously in comets. Figure 12 shows the detection of HDO and methanol in the comet Hyakutake.

Increasing amounts of data are being obtained at higher frequencies, i.e. in the submillimetre region. A recent survey of Orion was made between 607 and 725 GHz, and another one between 780 and 900 GHz started. These spectral regions are well suited for the detection of light hydrides. They are limited by the atmospheric windows. A continuous coverage will be available with the future satellite observatories, which are planned for the beginning of the third millennium.

II Radio Frequency Propagation