6.1.1 Transponder Assignment Modes

In addition to the multiple access techniques outlined in the preceding paragraphs, there are also certain transponder assignment modes. The commonly used ones include:

1. Preassigned multiple access (PAMA)
2. Demand assigned multiple access (DAMA)
3. Random multiple access (RMA)

In the case of preassigned multiple access (PAMA), the transponder is assigned to the individual user either permanently for the satellite's full lifetime or at least for long durations. The preassignment may be that of a certain frequency band, time slot or a code. When it is used infrequently, a link set-up with preassigned channels is not only costly to the user but the link utilization is also not optimum.

Demand assigned multiple access (DAMA) allows multiple users to share a common link wherein each user is only required to put up a request to the control station or agency when it requires the link to be used. The channel link is only completed as required and a channel frequency is assigned from the available frequencies within the transponder bandwidth. It is very cost effective for small users who have to pay for using the transponder capacity only for the time it was actually used. In the case of random multiple access (RMA), access to the link or the transponder is by contention. A user transmits the messages without knowing the status of messages from other users. Due to the random nature of transmissions, data from multiple users may collide.

In case a collision occurs, it is detected and the data are retransmitted. Retransmission is carried out with random time delays and sometimes may have to be done several times. In such a situation, when all the stations are entirely independent, there is every likelihood that the messages that collided would be separated out in time on retransmission.

6.2.1 Demand Assigned FDMA

In a demand assigned FDMA system, the transponder frequency is subdivided into a number of channels and the Earth station is assigned a channel depending upon its request to the control station. Demand assignment may be carried out either by using the polling method or by using the random access method. In the polling method, the master Earth station continuously polls all of the Earth stations in sequence and if the request is encountered, frequency slots are assigned to that Earth station which had made the request. The polling method introduces delays more so when the number of Earth stations is large. In the random access method, the problem of delays does not exist. The random access method can be of two types namely the centrally controlled random access method and the distributed control random access method. In the case of centrally controlled random access, the Earth stations make requests through the master Earth station as the need arises. In the case of distributed control random access, the control is exercised at each Earth station.

6.2.2 Pre-assigned FDMA

In a preassigned FDMA system, the frequency slots are pre-assigned to the Earth stations. The slot allocations are pre-determined and do not offer flexibility. Hence, some slots may be facing the problem of over-traffic, while other slots are sitting idle.

6.2.3 Calculation of C/N Ratio

The overall noise-to-carrier ratio for a satellite link is given by equation 6.1.

(6.1)

Where,

1. [N/C]_OV = Overall noise-to-carrier ratio
2. [N/C]_U = Uplink noise-to-carrier ratio
3. [N/C]_D = Downlink noise-to-carrier ratio
4. [N/C]_IM = Inter-modulation noise-to-carrier ratio

The value of noise-to-carrier ratio given by equation 6.1 should be less than the required design value of noise-to-carrier ratio [(N/C)_REQ], that is

(6.2)

Combining equations 6.1 and 6.2 we get,

(6.3)

In an FDMA system, the up-link noise is usually negligible and the inter-modulation noise is brought to an acceptable level by employing back-off in power amplifiers. It may be mentioned here that the operating point of the power amplifiers mostly TWTA is shifted closer to the linear portion of the curve in order to reduce the inter-modulation distortion. The reduction in the input power is referred to as input back-off and is the difference in dB between the carrier input at the operating point and the saturation input that is required for single carrier operation. The output back-off is the corresponding drop in the output power in dB. Therefore, equation 6.3 can be rewritten as

(6.4)

Equation 6.4 can again be rewritten as

(6.5)

The downlink carrier-to-noise ([C/N]_D) is expressed by equation 6.6.

(6.6)

Where,

1. [EIRP]_D = Satellite Equivalent Isotropic Radiated Power
2. [G/T]_D = Earth-station receiver G/T
3. [LOSSES]_D = Free space and other losses at the downlink frequency
4. [k] = Boltzmann's constant in dB
5. [B] = Signal bandwidth and is equal to the noise bandwidth

Therefore,

(6.7)

In the case of single carrier per channel (SCPC) systems, the satellite will have saturation value of EIRP ([EIRP]_sat) and transponder bandwidth of B_TR, both of which are fixed. In this case, there is no back-off and equality sign of equation 6.7 applies. Therefore,

(6.8)

Equation 6.8 can be rewritten as

(6.9)

Let us now consider the case of multiple carriers per channel (MCPC) systems having N carriers with each carrier sharing the output power equally and having a bandwidth of B. The output back-off is given by [BO]_O. The output power for each of the FDMA carriers is given by equation 6.10.

(6.10)

The transponder bandwidth (B_TR) is shared by all the carriers but due to power limitation imposed by the need of back-off, the whole bandwidth is not utilized. Let us assume that the fraction of the bandwidth utilized is alpha (α). Therefore,

(6.11)

Expressing in terms of decilogs

(6.12)

Substituting the values of [EIRP]_D and [B] given by equations 6.10 and 6.12 respectively in equation 6.7, we get

(6.13)

Equation 6.13 can be rearranged as

(6.14)

As we can see from equation 6.9, in the case of SCPC system LHS of equation 6.14 is zero. For MCPC systems, it is less than zero.

Therefore,

(6.15)

The best that can be achieved in a MCPC system is to make .

6.3.1 SCPC/FM/FDMA System

As outlined earlier, in this form of SCPC system, each signal channel modulates a separate RF carrier and the modulation system used here is frequency modulation. The modulated signal is then transmitted to the FDMA transponder. The transponder bandwidth is subdivided in such a way that each base band signal channel is allocated a separate transponder subdivision and an individual carrier. This type of SCPC system is particularly used on thin route satellite communication networks. Though it suffers from the problem of power limitation resulting from the use of multiple carriers and the associated intermodulation problems, it does enable a larger number of Earth stations to access and share the capacity of the transponder using smaller and more economic units as compared to multiple channels per carrier systems. Another advantage of the SCPC/FM/FDMA system is that it facilitates the use of voice activated carriers. This means that the carriers are switched off during the periods when there is no speech activity, thus reducing power consumption. This in turn leads to availability of more transponder power and hence higher channel capacity. This type of SCPC system also has the advantage that the power of the individual transmitted carriers can be adjusted to the optimum value for given link conditions. Some channels may operate at higher power levels than others, depending on the requirement of back-off for the transponder output power device. It may be mentioned here that the output back-off or simply the back-off of the transponder output power device is the ratio of the saturated output power to the desired output power. However, this type of SCPC system requires automatic frequency control to maintain spectrum centering for individual channels, which is usually achieved by transmitting a pilot tone in the centre of the transponder bandwidth.

Figure 6.3 shows the transmission path for an SCPC/FM/FDMA system. The diagram is self-explanatory. Different base band signals frequency-modulate their respective allocated carriers, which are combined and then transmitted to the satellite over the uplink.

The signal-to-noise power ratio (S/N) at the output of the demodulator for the SCPC/FM/ FDMA system can be computed from

(6.16)

where

1. C = carrier power at the receiver input (in W)
2. N = noise power (in W) in bandwidth B (in Hz)
3. B = RF bandwidth (in Hz)
4. f_d = test tone frequency deviation (in Hz)
5. f₂ = upper base band frequency (in Hz)
6. f₁ = lower base band frequency (in Hz)

6.3.2 SCPC/PSK/FDMA System

This is the digital form of the SCPC system in which the modulation technique used is phase shift keying (PSK). SPADE (single channel per carrier PCM multiple access demand assignment equipment) was the first operational SCPC/PSK/FDMA system. It was designed for use on Intelsat-4 and subsequent Intelsat satellites. This system employs PCM for base band signal encoding and QPSK as the carrier modulation technique. With this, it is possible to accommodate a 64 kbps voice channel in a bandwidth of 38.4 kHz as compared to the requirement of a full 45 kHz in the case of frequency modulation. With the use of 45 kHz per channel in QPSK, the guard band is effectively included in this bandwidth, which enables the SPADE system to handle 800 voice channels within a 36 MHz transponder bandwidth. The SPADE system offers the advantage of voice activation described earlier. The ECS-2 (European communications satellite) business service is another example of SCPC/PSK/FDMA system.

The channel capacity can be determined from the carrier-to-noise density ratio. The carrier-to-noise ratio necessary to support each carrier can be computed from

(6.17)

where

1. [C/N]_th = carrier-to-noise ratio(in dB)at the threshold error rate
2. [E_b/N₀]_th = bit energy-to-noise density ratio(in dB)at the threshold error rate
3. [B] = noise bandwidth(in Hz)
4. [R] = data rate(in bps)
5. [M] = system margin to allow for impairments(in dB)

The carrier-to-noise density ratio [C/N₀] can be computed from

(6.18)

where [C/N₀] is the total available carrier-to-noise density (in dB) for the transponder

6.4.1 MCPC/FDM/FM/FDMA System

In this arrangement, multiple base band signals are grouped together by using frequency division multiplexing to form FDM base band signals. The FDM base band assemblies frequency modulate pre-assigned carriers and are then transmitted to the satellite. The FDMA transponder receives multiple carriers, carries out frequency translation and then separates out individual carriers with the help of appropriate filters. Multiple carriers are then multiplexed and transmitted back to Earth over the downlink. The receiving station extracts the channels assigned to that station. Figure 6.4 shows the typical block schematic arrangement of such a system, which is suitable only for limited access use. The channel capacity falls with an increase in the number of carriers. Larger number of carriers causes more intermodulation products, with the result that intermodulation-prone frequency ranges cannot be used for traffic.

The signal-to-noise ratio at the demodulator output for such a system is given by:

(6.19)

where

1. f_d = RMS (root mean square) test tone deviation (in Hz)
2. f_m = highest modulation frequency (in Hz)
3. B = bandwidth of the modulated signal (in Hz)
4. b = base band signal bandwidth (in Hz)
5. C = carrier power at the receiver input (in W)
6. N = noise power (= kTB) in bandwidth B (in W)

6.4.2 MCPC/PCM-TDM/PSK/FDMA System

In this arrangement, multiple base band signals are first digitally encoded using the PCM technique and then grouped together to form a common base band assembly using time division multiplexing. This time division multiplexed bit stream then modulates a common RF carrier using phase shift keying as the carrier modulation technique. The modulated signal is then transmitted to the satellite, which uses FDMA to handle multiple carriers.

6.6.1 Reference Burst

The reference burst is usually a combination of two reference bursts (RB-1 and RB-2). The primary reference burst, which can be either RB-1 or RB-2, is transmitted by one of the stations, called the primary reference station, in the network. The secondary reference burst, which is RB-1 if the primary reference burst is RB-2 and RB-2 if the primary reference burst is RB-1, is transmitted by another station, called the secondary reference station, in the network. The reference burst automatically switches over to the secondary reference burst in the event of primary reference station's failure to provide reference burst to the TDMA network. The reference burst does not carry any traffic information and is used to provide timing references to various stations accessing the TDMA transponder.

6.6.2 Traffic Burst

Different stations accessing the satellite transponder may transmit one or more traffic bursts per TDMA frame and position them anywhere in the frame according to a burst time plan that coordinates traffic between various stations. The timing reference for the location of the traffic burst is taken from the time of occurrence of the primary reference burst. With this reference, a station can locate and then extract the traffic burst or portions of traffic bursts intended for it. The reference burst also provides timing references to the stations for transmitting their traffic bursts so as to ensure that they arrive at the satellite transponder within their designated positions in the TDMA frame.

6.6.3 Guard Time

Different bursts are separated from each other by a short guard time, which ensures that the bursts from different stations accessing the satellite transponder do not overlap. This guard time should be long enough to allow for differences in transmit timing inaccuracies and also for differences in range rate variations of the satellite.

6.7.1 Carrier and Clock Recovery Sequence

Different Earth stations have slight differences in frequency and bit rate. Therefore, the receiving stations must be able to establish accurately the frequency and bit rate of each burst. This is achieved with the help of carrier and clock recovery sequence bits. The length of this sequence usually depends on the carrier-to-noise ratio at the input of the demodulator and the carrier frequency uncertainty. A higher carrier-to-noise ratio and a lower carrier frequency uncertainty require a smaller bit sequence for carrier and clock recovery and vice versa.

6.7.2 Unique Word

Unique word is again a sequence of bits that follows the carrier and clock recovery sequence of bits in the preamble. In the reference burst, this bit sequence allows the Earth station to locate the position of the received TDMA frame. The unique word bit sequence in the traffic burst provides a timing reference on the occurrence of the traffic burst and also provides a timing marker to allow the Earth stations to extract their part of the traffic burst. The timing marker allows the identification of the start and finish of a message in the burst and helps to correct decoding. For obvious reasons, the unique word should have a high probability of detection. For instance, when the unique word of a traffic burst is missed, the entire traffic burst is lost. To achieve this, the unique word is a sequence of 1's and 0's selected to exhibit good correlation properties to enhance probability of detection.

Figure 6.7 shows a type of digital correlator circuit that can be used to detect the unique word bit sequence. Here, the unique word has N bits and is correlated with a stored pattern of itself. The received data is input to one of the N-bit shift registers, as shown in the figure, in synchronization with the data clock rate. The other N-bit shift register has a stored pattern of the unique word. Each stage of the shift register feeds a 2-bit adder, whose output is a ‘0’ if the bits are in agreement and a ‘1’ if they are in disagreement. The outputs of N adders are summed up. The output of the summer is a step function that depends upon the number of agreements or disagreements between the received unique word bit pattern and the stored pattern. The output of summer then feeds a threshold detector that specifies the acceptable number of disagreements. If the number of mismatches is less than or equal to the preset threshold value, the unique word is considered to have been detected. Remember that a declaration of the detection of the unique word occurs at the time instant of reception of the last bit or symbol of the unique word. As mentioned earlier, if the unique word belongs to the reference burst, it marks the TDMA receive frame timing. In the case of traffic burst, it marks the receive traffic burst timing.

6.7.3 Signalling Channel

The signalling channel is used to carry out system management and control functions. The signalling channel of the reference burst has three channels, namely (a) an order wire channel used to pass instructions to and from Earth stations, (b) a management channel transmitted by reference stations to all traffic stations carrying frame management instructions, such as changes in the burst time plan that coordinates traffic between different stations, and (c) a transmit timing channel that carries acquisition and synchronization information to different traffic stations, enabling them to adjust their transmit burst timing (TBT) so that similar bursts from different stations reach the satellite transponder within the correct time slot in the TDMA frame.

The signalling channel of the traffic burst also has an order wire channel, which performs the same function as it does in the case of reference burst. It also has a service channel, which performs functions like carrying traffic station's status to the reference station, carrying information such as the high bit error rate or unique word loss alarms, etc., to other traffic stations.

6.7.4 Traffic Information

Traffic bursts follow the reference burst in the TDMA frame structure. Each station in the TDMA network can transmit and receive many traffic bursts and sub-bursts per frame. The length of each sub-burst, which represents information on a certain channel, depends upon the type of service and the number of channels being supported in the traffic burst. For instance, while transmitting a PCM voice channel that is equivalent to a data rate of 64 kbps, each sub-burst for this channel would be 64 bits long if the frame time available for the purpose was 1 ms.

6.11.1 Extraction of Traffic Bursts from Receive Frames

To ensure a proper receive operation, the traffic station establishes the receive frame timing (RFT), which is defined as the time instant of occurrence of the last bit or symbol of the unique word of the primary reference burst. This sets the time marker from which the location of the traffic burst intended for a given station can be fixed. This is achieved by identifying the receive burst timing (RBT), which is determined by knowing the offset between the receive frame timing reference and the transmit burst timing. The amount offset in bits or symbols is contained in a receive burst time plan, which is stored in the foreground memory of the traffic station. Thus with the help of a receive burst time plan, the traffic station can extract all traffic bursts addressed to it in different frames. This whole process is called the receive frame acquisition.

6.11.2 Transmission of Traffic Bursts

A prerequisite for proper transmission of traffic burst is that it should reach the satellite transponder within the allocated position in the TDMA frame so as not to cause any overlap with traffic bursts transmitted to the transponder by other traffic stations. This can be ensured again by establishing what is called transmit frame timing (TFT) and transmit burst timing (TBT). While the former marks the start of the station's time frame, the latter marks the start of traffic burst. Again, the position of the traffic burst in a transmitted frame is determined by the offset between the transmit frame timing and the transmit burst timing. The information on the offset is contained in the transmit burst time plan stored in the foreground memory of the traffic station. The traffic burst that is transmitted at its transmit burst timing will fall into its appropriate position in the TDMA frame at the transponder. Similarly, traffic bursts from other stations accessing a particular transponder fall into their preassigned or designated positions in the TDMA frame without causing any burst overlapping. This whole process is called transmit frame acquisition.

6.11.3 Frame Synchronization

What has been discussed in the preceding paragraphs is the acquisition of receive frame and transmit frame timings. The acquisition process is needed when a station enters or re-enters operation. The process of maintaining the acquired timing references is synchronization. Synchronization becomes a necessity due to small changes in the satellite orbit caused by a variety of factors. A geostationary satellite orbit may be specified in terms of its inclination angle with the equatorial plane, its eccentricity and its east–west drift. A variation in the inclination angle causes north–south drift and a variation in eccentricity causes a variation in altitude. In the case of a geostationary orbit, the peak-to-peak variation in altitude is 0.2 % of the orbit radius (= 42 164 km), which amounts to about 85 km (0.2 × 42164/100). Peak-to-peak north–south drift and peak-to-peak east–west drift are about 0.2°, which amounts to about 150 km in terms of distance. These errors introduce a maximum range variation of 172 km [= √ (85² + 150²)]. This is equivalent to a one-way propagation delay of 0.575 ms and a round trip maximum delay variation of 1.15 ms. Assuming that the satellite takes about 8 hours to move from its nominal position to a position where maximum delay variation occurs, this leads to a maximum Doppler shift of 40 ns/s. This Doppler shift causes errors in the burst positions as they arrive at the transponder. Thus, frame synchronization is necessary for proper transmission and reception of traffic bursts.

6.12.1 Advantages of TDMA over FDMA

1. The bit capacity of the TDMA system is independent of the number of accesses.
2. The power amplifiers in the transponder of a TDMA system can be operated in the saturation mode as opposed to that of an FDMA system. This increases the capacity of multiple accesses.
3. The duty cycle of the Earth station in a TDMA system is low.
4. TDMA systems allow the use of digital techniques like digital speech interpolation, satellite on-board switching and so on.
5. The TDMA systems offer more flexibility due to use of high-speed logic circuits and processors which offer high data rates.
6. TDMA systems are more economical as compared to FDMA systems as they are easy to multiplex, independent of distance and can be easily interfaced with terrestrial services.
7. TDMA systems can tolerate higher levels of interference noise.

6.12.2 Disadvantages of TDMA over FDMA

1. The peak power of the amplifier in a TDMA system is always large.
2. The TDMA system is more complex as compared to the FDMA system. The Earth station of the TDMA system requires ADC, clock recovery, synchronization, burst control and data processing before transmission.
3. TDMA systems use high-speed PSK/FSK circuits.

6.13.1 DS-CDMA Transmission and Reception

Figure 6.11 shows the basic DS-CDMA transmitter. The diagram is self-explanatory. The transmitter generates a bit stream by multiplying in the time domain the message bit stream m_i(t) and the code information a_i(t). Multiplication in the time domain is convolution in the frequency domain. Therefore, the product of m_i(t) and a_i(t) produces a signal whose spectrum is nothing but convolution of the spectrum of m_i(t) and the spectrum of a_i(t). Also, if the bandwidth of the message signal is much smaller than the bandwidth of the code signal, the product signal has a bandwidth approaching that of the code signal.

The message signal is either an analogue or a digital signal. In the majority of cases, it is a digital signal. In that case, message signal modulation as shown in the diagram is omitted and the message signal is directly multiplied by the code signal. The resulting signal then modulates a wideband carrier using a digital modulation technique, which is usually some form of phase shift keying.

Figure 6.12 shows the basic block schematic arrangement of the DS-CDMA receiver. The diagram is self-explanatory. It is assumed that the receiver is configured to receive the message signal m_i(t). The receiver in this case generates a code signal a_i(t) synchronized with the received message. If the signals represented by suffix j constitute undesired signals, i.e. noise, then the bit stream present at the output of the first stage of the receiver and at the input of the demodulator is given by

(6.27)

In the case of orthogonal codes,

(6.28)

(6.29)

By substituting these values, it is found that only the first part of the expression for the received bit sequence is the desired one; the remaining part is just noise.

A term called processing gain (G) is defined in such systems. It is given by

(6.30)

A chip rate of 2.5 Mbps and a message bit rate of 25 kbps would give a processing gain of 20 dB. Since the undesired component or the noise is spread over the entire bandwidth and the receiver responds only to the 1/G part of this bandwidth, this has the effect of reducing the noise by the same factor. The thermal noise and intermodulation noise are also reduced by the factor G. To sum up, the input to the DS-CDMA receiver is wide band and contains both desired and undesired components. When this received bit stream is applied to the correlator, the output of the correlator is the desired message signal centred on the intermediate frequency. The undesired signals remain spread over the entire bandwidth and only that portion within the receiver bandwidth causes interference. This is further illustrated in Figure 6.13.

In practice, the cross-correlation function is not equal to zero, which leads to some performance degradation. The level of interference is directly determined by the ratio of the peak cross-correlation value to the peak auto-correlation value of the pseudorandom code sequences.

DS-CDMA is further divided into two types, namely:

1. Sequence synchronous DS-CDMA
2. Sequence asynchronous DS-CDMA

6.13.2 Frequency Hopping CDMA (FH-CDMA) System

CDMA is also referred to as spread spectrum multiple access because of the reason that the carrier spectrum is spread over a much larger bandwidth as compared to the information rate. The spread spectrum signal is inherently immune to jamming as it forces the jammer to deploy its transmitted jamming power over a much wider bandwidth than would have been necessary for a conventional system. In other words, for a given power of the jammer, the power spectral density produced by the jammer becomes reduced by a factor of increase in the bandwidth due to spreading. The direct-sequence CDMA discussed in earlier paragraphs is one way to spread the carrier spectrum where the message bit sequence is multiplied by a pseudorandom code sequence. Frequency hopping (FH), to be discussed in the following paragraphs, is another method used to do the same.

In the case of a frequency hopping CDMA (FH-CDMA) system, the carrier is sequentially hopped into a series of frequency slots spread over the entire bandwidth of the satellite transponder. The transmitter operates in synchronization with the receiver, which remains always tuned to the frequency of the transmitter. The transmitter transmits a short burst of data on a narrowband, then tunes to another frequency and transmits again. The transmitter thus hops its frequency over a given bandwidth several times per second, transmitting one frequency for a certain period of time, then hopping to another frequency and transmitting again. This is achieved by using a frequency synthesizer whose output is controlled by a pseudorandom code sequence. The pseudorandom code sequence decides the instantaneous transmission frequency. On the receiver side, the data can be recovered by using an identical frequency synthesizer controlled by an identical pseudorandom sequence. Figures 6.14 and 6.15 show the block schematic arrangements of typical FH-CDMA transmitter and receiver respectively. The diagrams are self-explanatory.

Due to the random hopping pattern as governed by the pseudorandom sequence, to an observer, the carrier appears to use the entire transponder bandwidth over the pseudorandom sequence period. At a given instant of time, however, it uses a particular frequency slot. The hopping rate of the carrier may equal the coded symbol rate in the case of slow hop systems or be several times that of the coded symbol rate in the case of fast hop systems. In the case of FH-CDMA, each Earth station is assigned a unique hop pattern. Commonly used modulation schemes in frequency hop CDMA systems include noncoherent M-ary FSK (FSK techniques having M frequency levels) and differential QPSK. Also, non-coherent demodulation is used as it is very difficult to maintain phase coherence between the hops.

6.13.3 Time Hopping CDMA (TH-CDMA) System

In the case of the time hopping CDMA (TH-CDMA) system, the pseudorandom bit sequence determines the time instant of transmission of information. In fact, the signal is transmitted by the user in rapid bursts during time intervals determined by the pseudorandom code assigned to the user. A given user transmits only during one of the M time slots each frame has been divided into. However, the time slot used by a given user for transmission of data in successive frames depends upon the code assigned to it. Since each user transmits its data only during one of the M time slots in each frame, the bandwidth available to it increases by a factor of M. Figure 6.16 shows the block schematic arrangement of a typical TH-CDMA transmitter. The typical receiver block schematic is shown in Figure 6.17.

6.13.4 Comparison of DS-CDMA, FH-CDMA and TH-CDMA Systems

The operational principle of the three systems can be illustrated with the help of frequency–time graphs, as shown in Figures 6.18 (a), (b) and (c). These graphs depict the frequency usage of the three systems as a function of time. It is evident from the graph shown in Figure 6.18 (a) that a DS-CDMA system occupies the whole of the available bandwidth when it transmits, whereas an FH-CDMA system uses only a small part of the bandwidth at a given instant of time when it transmits, as shown in Figure 6.18 (b). However, the location of this part differs with time. The bandwidth occupied by an FH-CMDA signal for a given hop frequency depends not only on the bandwidth of the message signal but also on the shape of the hopping signal and the hopping frequency. In the case of a slow hop system, when the hopping frequency is much smaller than the message signal bandwidth, the occupied bandwidth is mainly decided by the message signal bandwidth. In case of a fast hop system, when the hopping frequency is much larger than the message signal bandwidth, the occupied bandwidth is mainly decided by the shape of the hopping signal at a given hopping frequency.

In the case of time hopping CDMA (TH-CDMA), depicted in Figure 6.18 (c), as compared to FH-CDMA, the whole of the available bandwidth is used for short time periods instead of parts of the bandwidth being used all the time. To sum up, DS-CDMA uses an entire bandwidth all the time, FH-CDMA uses a small part of the bandwidth at a given time instant but the chosen frequency slot varies with time in order to cover the entire bandwidth and TH-CDMA uses the entire bandwidth for short periods of time.

6.14.1 Frequency Re-use in SDMA

Frequency re-use, as outlined above, is the key feature and the underlying concept of space domain multiple access (SDMA). In the face of continually increasing demands on the frequency spectrum, it becomes important that frequency bands assigned to satellite communications are efficiently utilized. One of the ways of achieving this is to re-use all or part of the frequency band available for the purpose. Another way could be employment of efficient user access methods. Yet another approach could be the use of efficient modulation, encoding and compression techniques in order to pack more information into available bandwidths.

Restricting the discussion to frequency re-use, which is the present topic, the two methods in common use today for the purpose are beam separation and beam polarization. Beam separation is based on the fact that if two beams are so shaped that they illuminate two different regions on the surface of the Earth without overlapping, then the same frequency band could be used for the two without causing any mutual interference. One could do so by using two different antennas [Figure 6.19(a)] or a single antenna with two feeds [Figure 6.19(b)].

Beam polarization, on the other hand, relies on the principle of using two orthogonally polarized electromagnetic waves to transmit and receive using the same frequency band with no mutual interference between the two. Orthogonal polarizations used commonly include horizontal and vertical polarizations or right-hand circular and left-hand circular polarizations.

Both techniques have the capability of doubling the transmission capacity individually, and when used in tandem can increase the capacity four times. SDMA is seldom used in isolation. It is usually used in conjunction with other types of multiple access techniques discussed earlier, including FDMA, TDMA and CDMA. In the following paragraphs, the employment of SDMA separately with each one of these multiple access techniques will be briefly discussed.

6.14.2 SDMA/FDMA System

Figure 6.20 shows a typical block schematic arrangement of the SDMA/FDMA system in which the satellite uses fixed links to route an incoming uplink signal as received by a receiving antenna to a particular downlink transmitter antenna. It is clear from the diagram that the satellite uses multiple antennas to produce multiple beams. The transmitting antenna–receiving antenna combination defines the source and destination Earth stations.

The desired fixed links can be set on board the satellite by using some form of a switch, which can be selected only occasionally when the satellite needs to be reconfigured. The links could also be configured alternatively by switching the filters with a switch matrix operated by a command link. It may once again be mentioned here that satellite switches are changed only occasionally when the satellite is to be reconfigured.

6.14.3 SDMA/TDMA System

This system uses a switching matrix to form uplink/downlink beam pairs. In conjunction with TDMA, the system allows TDMA traffic from the uplink beams to be switched to the downlink beams during the course of a TDMA frame. The link between a certain source–destination combination exists at a specified time for the burst duration within the TDMA frame. As an example, the signal on beam 1 may be routed to beam 3 during, say, the first 40 μs of a 2 ms TDMA frame and then routed to beam n during the next 40 μs slot. The process continues until every connection for the traffic pattern has been completed. Figure 6.21 shows a typical transponder arrangement for an SDMA/SS/TDMA system. (SS here stands for ‘satellite switched').

6.14.4 SDMA/CDMA System

CDMA provides multiple access to the satellite. The satellite receives an uplink CDMA bit stream, decodes it to determine the destination address and then routes it to the desired downlink. The bit stream is usually re-timed, regenerated and stored in onboard processors before it is retransmitted. This implies that a downlink configuration does not need to be the same as the uplink configuration, thus allowing each link to be optimized.

Internet Sites

1. http://www.bee.net/mhendry/vrml/library/cdma/cdma.htm
2. http://www.iec.org/online/tutorials/tdma/index.html
3. www.sergiochacon.com/Communication/Wireless/Satellite TDMAFDMA.ppt
4. http://www.mlab.t.u-tokyo.ac.jp/ mori/courses/WirelessMobileComm2000/CDMA.pdf
5. http://www.eas.asu.edu/ junshan/pub/net_capacity.pdf
6. http://www.umtsworld.com/technology/cdmabasics.htm
7. http://en.wikipedia.org/wiki/Space-division_multiple_access