Microwave Communication Systems

Chapter 19 Microwave Communication Systems Abstract This chapter presents terrestrial mobile communication and satellite communication systems with emphasis on design and analysis of various subsys- tems used in receiver and transmitter sections. Homodyne and heterodyne receivers and different transmitter architectures for wireless mobile communication systems are described. Satellite communication systems defining important parameters like EIRP, G/T and SFD are also presented. Detailed design, analysis and realization of various microwave subsystems for satellite transponders like receivers, channel amplifiers, linearizers, TWTAs, SSPAs, microwave power module (MPM) and mul- tiport power amplifier (MPA) are also described. 19.1 Introduction Wireless communication technology is the fastest growing technology in the world. The successful development of wireless technologies has greatly improved people’s ability to remain connected socially as well as for business. The phenomenal success of wireless mobile communications is reflected by the rapid pace of technology inno- vation. From the first-generation (1G) analog mobile communication system in 1980 to the 3G system which was first launched in 2001 through the second-generation (2G) mobile communication system in 1991, the wireless mobile network has trans- formed from a pure telephony system to a rich multimedia content transport system. The fourth-generation (4G) wireless system can support data rates of up to 1 Gbps for low mobility and up to 100 Mbps for high mobility. The target of a wireless communication system is to connect and achieve seamless communications between anybody, anything, wherever they are, whenever they need and by whatever elec- tronic means they wish with affordable cost by consuming less power. Upcoming wireless communication systems are aiming to fulfil these requirements by configur- ing next-generation terrestrial wireless and satellite communication systems. As the lower frequency bands become congested, the communication systems need to shift towards higher and higher frequency bands to meet the capacity and data rate targets. Higher microwave and millimetre wave frequency bands are much less congested © Springer Nature Singapore Pte Ltd. 2019 583 S. C. Bera, Microwave Active Devices and Circuits for Communication, Lecture Notes in Electrical Engineering 533, https://doi.org/10.1007/978-981-13-3004-9_19 584 19 Microwave Communication Systems and can potentially provide multi-gigahertz spectrum. This is the reason why the satellite communication systems are already using higher microwave and mmwave frequency bands for high-throughput applications. It is also envisaged that the next- generation (5G) mobile communication system needs to use higher microwave and mmwave frequency bands to meet the requirement of spectrum [1]. This chapter will discuss terrestrial mobile communication and satellite communication systems with associated microwave circuits and subsystem. 19.2 Mobile Communication Systems Mobile communication system is one of the greatest innovations in the history of technology. The rapid adoption of smartphones by a wide range of users and the explosive growth of data traffic of these devices have been phenomenal. A simplified block diagram of a generic transceiver for wireless communication system is shown in Fig. 19.1. It consists of transmit and receive RF section and a baseband processing section. The transmitter modulates the carrier signal with particular scheme and transmits through an antenna. The receiver recovers the information from the signal received by the antenna. A bandpass signal v(t) can be considered as a sinusoidal waveform of frequency fc , with varying amplitude A(t) and/or phase angle ϕ(t) with time, which can be written as   v(t)  A(t) cos 2π fc t + ϕ(t) (19.1) Suppose vi (t)  A(t) cos ϕ(t) (19.2a) Fig. 19.1 Block diagram of generic wireless communication system 19.2 Mobile Communication Systems 585 vq (t)  A(t) cos(ϕ(t) + 90◦ ) (19.2b) Then, (19.1) can be expressed as v(t)  vi (t) cos(2π fc t) + vq (t) cos(2π fc t + 90◦ ) (19.3) This shows that a bandpass signal can be expressed completely by its in-phase and quadrature-phase components. The components vi (t) and vq (t), respectively, are the in-phase (I) and quadrature-phase (Q) components of the signal. In a generic transmitter, the baseband signal processor converts the information signal into an in- phase (I) and quadrature-phase (Q) component in the digital domain. The properties of the I–Q signals determine the type of modulation. Thus, the actual data symbols to be transmitted are coded to the I and Q signals in baseband circuitry. The I–Q signals are finally amplified by the power amplifier and transmitted through an antenna. In the generic receiver, the received signal from the antenna is amplified by a low-noise amplifier and finally the transmitted symbols are decoded by recovering the original I and Q signals from the received RF signal. The amplitude and phase information of the signal can be extracted from the I–Q signals as follows:  A(t)  vi (t)2 + vq (t)2   vq (t) ϕ(t)  tan−1 vi (t) 19.2.1 Receiver Architecture for Mobile Communication The basic function of a wireless communication receiver is to distinguish the desired signal from all other signals and to amplify the signal to a sufficient level for reli- able detection by its baseband signal processing section. The simplest topology of its RF section is a bandpass filter tuned to the same frequency with the signal. The frequency response of a band select filter receiver in the presence of adjacent chan- nels/interferences is shown in Fig. 19.2. The filter should be a tunable or there should be a bank of selectable bandpass filters with different centre frequencies to select one channel at a time among desired multiple channels. This scheme is suitable only if desired channels are separated sufficiently, and there should not be any inter- ference signal near to the desired signals to design a practical filter to select only one channel in the presence of other channels. However, in practice, many narrow frequency bands spaced very closely to utilize the total allocated frequency bands with minimum wastage. Thus, it becomes very difficult to design a practical filter to select only the desired channel suppressing all other closely spaced channels and interferences. In some situations, the signal strength of the neighbouring channels becomes much higher than the selected channel due to the path loss and/or channel fading. In a practical scenario, a channel filter with very sharp cut-off response is 586 19 Microwave Communication Systems Fig. 19.2 Band-select filter receiver response required to select a narrow bandwidth signal at higher carrier frequency. This needs a filter with very high-quality factor which is practically impossible at least in an inte- grated form. To overcome this problem practically, heterodyne as well as homodyne receiver architectures are used for wireless communication receivers. In heterodyne receiver architecture, the requirement of an extremely narrowband channel select filter centred around a higher carrier frequency is eliminated by per- forming the channel selection at a lower frequency. Block diagram of simple hetero- dyne receiver architecture is shown in Fig. 19.3. In this case, the channel select filter operates at a lower frequency with unchanged pass bandwidth leading to practical realizable filter with achievable quality factor. Another advantage of this architecture is that the different channels can be selected using fixed channel select filter sim- ply by changing the frequency of the local oscillator. The frequency translation is achieved by a process called heterodyning. In heterodyning, the original signal vi (t) at frequency fo1 is translated to a new intermediate frequency (IF) fo1 − fLO by mul- tiplying it in the time domain with another signal vLO (t) at frequency fLO . Cascading several mixing stages with an appropriate filter, the signal spectrum can be translated gradually to any other frequency band. Such receivers are called superheterodyne receivers. Though the heterodyne receiver architecture solves the problem of channel selection, it suffers from the requirement of image reject filter and associated inser- tion loss of it. There are some other receiver architectures to overcome the image rejection problem such as image reject mixer, direct conversion, low IF and wideband IF. Among them, image reject mixers are popularly used and already discussed in the previous chapter. The direct-conversion receiver is also known as zero-IF receiver and homodyne receiver. Block diagram of zero-IF receiver architecture is shown in Fig. 19.4. Here, the centre frequency of the incoming signal and the LO signal are same, and thus the mixer translates the input spectrum directly around DC. The low-pass filters at outputs of the mixers attenuate all unwanted signals outside the desired frequency band and divide the signal into I and Q components for further processing by 19.2 Mobile Communication Systems 587 Fig. 19.3 Block diagram of heterodyne receiver Fig. 19.4 Block diagram of a homodyne receiver digital processor for information extraction. The homodyne receivers have several advantages over the conventional heterodyne architecture. These architectures do not require any highly selective image reject filters; this leads to savings in size. Here, simple low-pass filters are designed at lower frequencies as an integrated unit to select the desired channel. The homodyne architecture is attractive for the possibility of realizing integrating receiver using low-frequency amplifiers and digital-to-analog converter (DAC) followed by digital signal processing unit just after the low-pass filters. However, there are some practical issues to implement the homodyne receivers since the desired signal translated around DC. The architecture is highly sensitive to all sources of DC power, such as LO self-mixing, re-receiving of LO signal after reflection from other circuitry (antenna), DC offsets in the circuitry, down-mixing of signals due to even-order distortion in LNA and mixer, etc. This architecture also sufferers from 1/f noise of active devices working in the low-frequency regions. Thus, to achieve full benefit from the homodyne architecture, both the LNA and the mixers should be highly linear to avoid even-order distortion. The low IF architecture receiver is the combination of good properties of the image reject receiver and the zero-IF architecture. In this architecture, the intermediate 588 19 Microwave Communication Systems frequency (IF) is non-zero, but it is much lower than the conventional image reject receiver. The low IF frequency leads to the use of low-frequency/low-power circuit configuration which is more suitable for integration than the structures of the image reject receiver. It also helps to circumvent the DC offset and 1/f noise problems of the zero-IF architecture. 19.2.2 Transmitter Architecture for Mobile Communication The function of a wireless communication transmitter is to modulate a carrier frequency by information signal with particular scheme and to transmit through an antenna. Contrary to the receiver, the transmitter circuitry processes signals of well-known strength and spectral contents. The challenges in transmitter design are mainly related to the spectral purity of the final power-amplified modulated signal entering the antenna, particularly in a system where many channels are closely spaced in frequency domain. Practically, the final power amplifier which amplifies the modulated signals generally operates in nonlinear region to achieve better DC-to-RF efficiency. Thus, the most severe source of spectral deterioration is due to nonlinear- ity of the power amplifier that generates spectral re-growth which disturbs in-band as well as adjacent channel signals. The power amplifier is generally designed using III–V group semiconductors for achieving high RF power level with better efficiency and it is integrated with linearizer to minimize the overall nonlinear effect. Block diagram of a simple architecture direct-conversion transmitter is shown in Fig. 19.5. The direct-conversion transmitter consists of an I/Q modulator, bandpass filter and power amplifier. The simple architecture suffers severely from sensitivity to the imbalance in between the mixers and to the phase and amplitude errors between the LO signals. These error mechanisms will lead to increased LO leakage and non-optimal suppression of the image frequency. Another shortcoming of the direct- conversion transmitter is the sensitivity to LO pulling, where the strong output signal of the power amplifier disturbs the operation of local oscillator due to their same frequency. This effect becomes worse if power amplifier and local oscillator are integrated into the same chip. One way to minimize LO pulling is to offset the local oscillator from the output frequency. Two local oscillators with offset frequency may be used to produce final frequency using a mixer. In this case, it is possible to keep frequency of the LOs far away from the output frequency. Another way to reduce the LO pulling is the use of heterodyne transmitter architecture as shown in Fig. 19.6. Here, the I/Q modulation is performed at a lower LO frequency and the final frequency translation is done by a second mixer stage to translate the spectrum to output frequency. 19.2 Mobile Communication Systems 589 Fig. 19.5 Block diagram of a direct-conversion transmitter Fig. 19.6 Block diagram of a heterodyne transmitter 19.2.3 Transceiver for Mobile Communication Though the functionalities of the transmitter and receiver for base station and for the mobile units are same, the practical realizations are different due to the different driving forces behind their development. The integration of the mobile transmitter and receiver units is extremely desirable to extend battery life, reducing weight and size which make the handheld mobile unit more desirable and attractive to the customer. This can be done by increasing the level of integration. Reliability of the integrated complete unit is also directly influenced by the integration of multiple functions into a single chip and the mass production of ICs. The RF section of a transmitter usually consists of an oscillator, a modulator, an upconverter, filters and power amplifier as shown in Fig. 19.7. It also consists of a phase-locked oscillator or synthesizer. The signals to be transmitted modulate 590 19 Microwave Communication Systems Fig. 19.7 Transceiver block diagram for mobile communication systems the oscillator frequency through a suitable modulation scheme such as amplitude modulation, frequency modulation, phase modulation or other digital modulation schemes such as PSK, FSK, etc. In WCDMA system, QPSK modulation scheme is used as shown in the Fig. 19.7. Here, the data stream from baseband processor is split into odd and even streams and are separately modulated before combining into a QPSK signal. The signals are then amplified by amplifier and upconverted to a higher desired carrier frequency by an upconverter. The upconverted RF signals are amplified by a power amplifier and transmitted by the antenna. The function of the receiver is to extract the useful information from the received RF signals from antenna. Receiver section in an RF transceiver consists of a low-noise amplifier, downconverter, IF amplifier, bandpass filters and demodulator. The demodulated signal processed in the baseband processor for extraction of the information. Enormous growth of data traffic volume is one of the main drivers behind the next- generation (5G) mobile communication systems [1]. Mobile wireless communication with ultra-wide bandwidth will be the key requirement for the next-generation mobile system to meet the ever-increasing needs for higher capacity. Presently, wireless communication demands tens of billions of Internet of things (IoT) and machine- to-machine (M2M) communications. The next-generation wireless communication system needs enormous spectrum to fulfil the requirements for smart city, smart home, smart grid, smart remotely controlled and self-driven vehicle, e-health which is available only at higher microwave and millimetre wave frequency bands. In addi- tion to the mmwave frequency bands, the fundamental technologies that enable next- generation mobile communications include massive multiple-input-multiple-output (MIMO) system and small-cell configuration. The mmwave technologies enable to utilize the wideband spectrum; massive MIMO extends the range and increases spec- tral efficiency in these frequencies by employing a large number of antennas, and 19.2 Mobile Communication Systems 591 small-cell technologies provide the means to deploy a wide area mobile network with a large number of small cells, scalable backhaul and proper interference mitigation techniques. All these technologies together enable next-generation mobile commu- nication systems with much larger capacity, much higher data rate and much denser deployment than 4G systems. It is envisaged that the next-generation mobile com- munication systems (5G) will work in higher microwave and mmwave frequency bands. The main functions of the RF section of the transceiver chain are amplification of signal, conversion of frequencies and analog beamforming. The signal amplifi- cation is done by power amplifiers in transmit chain and by LNA in receive chain. The frequency conversions are done by the mixers using suitable local oscillators as shown in Fig. 19.8. At higher microwave and mmwave frequencies, patch antenna arrays are most viable option for implementing antenna arrays with large number of elements in a small size to achieve sufficient range. A simplified block schematic of a mobile transceiver with hybrid beamforming architecture using 64 patch elements is shown in Fig. 19.8. Maintaining most of the capability to dynamically form beams and adapt the MIMO processing schemes according to the channel conditions, it is required to minimize the number of RF chains for reduction of cost and complexity. Here, this is shown by using antenna sub-arrays, analog beamforming and digital MIMO processing. Analog beamforming can be implemented using RF, IF or LO beamforming technology. Here, it is shown in RF section. The 64 antenna elements are grouped into 16 numbers of 4 × 1 sub-arrays, and thus only 16 numbers of RF chains are required for 64 elements. Each RF signal is phase shifted and combined in a group of four before frequency conversion. This reduces the number of con- verter and MIMO streams. This scheme needs to handle only four MIMO streams with four converters (each in transmit and receive section) to drive the entire 64- element antenna array which reduces the complexity in the digital processing unit. For RF signal phase shifting, reflective-type phase shifter, loaded line phase shifter or switched delay phase shifters can be implemented at microwave and mmwave frequency range. RF section of the mobile transceiver mainly consists of low-noise amplifier (LNA), power amplifier (PA), up- and downconverters, IF amplifiers, phase shifters, modula- tors, switches and filters as shown in Fig. 19.8. There are several viable semiconduc- tor device technologies to realize RF transceiver circuits working at microwave and mmwave frequency bands. Gallium arsenide (GaAs) and indium phosphide (InP) and high-electron-mobility transistor (HEMT) and hetero-junction bipolar transistor (HBT) devices based on GaAs and SiGe can be used for solid-state power amplifiers to transmit few Watts of power. GaAs and InP p-HEMTs are among the best choices for RF Tx/Rx switches and LNAs. SiGe HBTs and RF CMOS have also been used for low-cost transceiver circuits in low-to-medium RF power amplifications. The mobile transceiver can be realized in multi-chip modules and surface mount packages using MMIC chips. Thus, the whole transceiver can be integrated on a single printed circuit board. 592 19 Microwave Communication Systems Fig. 19.8 Block diagram of mobile transceiver with MIMO system 19.2 Mobile Communication Systems 593 Another critical technology required for the 5G mobile communication is the high-speed analog-to-digital converter (ADC). To support multi-Gbps data rate, it is required to use multiple channels of ADC with a sampling rate of few hundred MHz to few GHz. 19.3 Satellite Communication Systems Microwave frequencies are used in satellite communication due to its higher pene- tration capability through the atmosphere, higher percentage bandwidth and higher directive coverage. A microwave communication system mainly consists of a receiver, a transmitter and an antenna system. Block diagram of a satellite com- munication system is shown in Fig. 19.9. Like any other communication systems, the space and the ground segments consist of a transmitter, a receiver and an antenna system. A simple transponder for satellite communication is a repeater system. It amplifies the signal received by a receive antenna with minimum addition of noise and then translates the signal frequency band to other bands. The frequency-translated signal further amplified to high power level by transmitter and then transmitted to ground by a transmit antenna. For example, Fig. 19.9 shows a C × S-band transpon- der. It receives the signal in C-band and transmits it in S-band. Frequency translation provides high level of isolation to the sensitive receive sections from the high level of transmitted RF power. This ensures the stability of the transponder avoiding reinjec- tion of transmitted output signal to the receive section by using filtering. A satellite transponder has three most important performance parameters: saturation flux den- sity (SFD), receive antenna gain (GR ) to system temperature (TR ) ratio (GR /TR ) and satellite-saturated effective isotropic radiated power (EIRP). Saturation flux density is the received power flux density at the satellite which saturates the transponder, i.e. to saturate final high power amplifier (HPA) used in the transmitter. Practically, the satellite SFD is not an independent parameter; its minimum value is limited by the ratio GR /TR . Suppose the equivalent noise temperature of a satellite receiver (including losses in between receiver and receive antenna output) is Te , and the total equivalent system noise power Pn can be written as Pn  k(Te + To )B (19.1a) Here, B is the noise bandwidth of the system and To is the antenna noise temper- ature which is about 290 K and k is Boltzmann constant. Thus, the overall system noise temperature TR is (Te + To ). The received power PR at the output of the satel- lite receive antenna should be more than the total noise power Pn . Considering the received power flux density at the satellite is PFD, the received power at the output of the receive antenna can be written as PR  PFD × Ae 594 19 Microwave Communication Systems Fig. 19.9 A simplified block diagram of a satellite communication system Here, Ae is the effective aperture of the receive antenna, and it is related to the gain of the antenna GR by GR  4π Ae /λ2R Here, λR is the wavelength corresponding to the receive signal frequency. Thus, the received power is given by PFD × GR × λ2R PR  (19.2) 4π For reliable communication, the received power must be more than the total noise power. The amount will be determined by the required signal-to-noise power ratio needed for particular data rate, used modulation code, etc. for a communication. Here, to derive the expression for minimum SFD level, we will impose the condition PR > Pn (19.3) Combining (19.1), (19.2) and (19.3), PFD × GR × λ2R > kTR B (19.4) 4π or 4π TR PFD > ×k ×B× λR 2 GR Expressing all the parameters in dB (19.4), it can be written as 19.3 Satellite Communication Systems 595     4π GR [PFD]dBW/m2 > + [k]dBW/Hz−K + [B]dBHz − (19.5) λ2R dB/m2 TR dB/K This shows that for a transponder operating at particular frequency and bandwidth, the minimum operating PFD limit is determined by its receive antenna gain-to-system noise temperature ratio, i.e. (GR /TR ). Depending upon the requirement of uplink carrier-to-noise power ratio, which is determined by the communication data rate, modulation codes, etc.; the operating PFD limit is selected which is higher than the value given by (19.5). Example 19.1 A satellite transponder operates at receive frequency 5.9 GHz of band- width 35 MHz. The gain of the receive antenna is 27 dBi and the receiver noise figure (including loss in between receive antenna and receiver) is 3.5 dB. Calculate the min- imum received power flux density limit of the transponder. Solution Receive frequency  5.9 GHz Thus, λR  (0.3/5.9) m  0.05085 m Receiver noise figure NF  3.5 dB  2.2387(in factor) Receiver noise temperature Te  (NF − 1)To  (2.2387 − 1)290 K  359 K Thus, the system noise temperature TR  Te + To  (290 + 359) K  649 K  28 dBK Thus,   GR  (27 − 28) dB/K  −1 dB/K TR dB/K     4π 4π  10 log  36.87 dB/m2 λR dB 2 λ2R [k]dBW  −228.6 dBW/Hz − K [B]dBHz  10 log(35 × 106) dB − Hz  75.4 dBHz     4π GR [PFD]dBW/m2 > + [k]dBW/Hz−K + [B]dBHz − λ2R dB/m2 TR dB/K   GR > 36.87 − 228.6 + 75.4− TR dB/K 596 19 Microwave Communication Systems   GR > −116− TR dB/K > −115 dBW/m2 Thus, the minimum PFD is −115 dBW/m2 . The saturated power flux density (SFD) also should be higher than the value given by (19.5). Otherwise, the transponder will be saturated with the noise power itself. Saturated effective isotropic radiated power (saturated EIRP) of a satellite transponder is determined by the final high power amplifiers output power capa- bility (Psat ), output loss (LOUT ) and satellite’s transmit antenna gain (GT ). The output loss is the RF loss of all the elements in between output of the power amplifier and input of the antenna. Therefore, saturated RF power at the input of the antenna (PT sat ) is given by PT sat  Psat − LOUT (19.6a) Thus, the saturated EIRP is given by EIRPsat  PT sat × GT (19.6b) The overall gain of a transponder is determined by its saturated EIRP and SFD. From (19.2), under saturated flux density condition, the received power at the input of the receiver is given by SFD × GR × λ2R PRsat  (19.7) 4π Therefore, the overall gain (GTransponder ) of the transponder is given by PT sat EIRPsat GTransponder   (19.8) PRsat GT × PRsat   GTransponder dB  [EIRPsat ]dBW − [GT ]dBi − [PRsat ]dBW (19.9) Example 19.2 A satellite transponder operates at receive frequency 5.9 GHz and transmit frequency 2.6 GHz of bandwidth 35 MHz. The gain of receive and transmit antennae are 27 and 42 dBi, respectively. Calculate the gain of the transponder if SFD is −95 dBW/m2 and saturated EIRP is 65 dBW. Solution Receive frequency  5.9 GHz Thus, λR  (0.3/5.9) m  0.05085 m Receive antenna gain  27 dBi From (19.6b), the saturated transmit power 19.3 Satellite Communication Systems 597 EIRPsat PT sat  GT [PT sat ]dBW  [EIRPsat ]dBW − [GT ]dBi  65 dBW − 42 dBi  23 dBW  53 dBm From (19.7), SFD × GR × λ2R PRsat  4π   λ2R [PRsat ]dBW  [SFD]dBW/m2 + [GR ]dBi + 4π dBm2  (−95 + 27 − 36.87) dBW  −104.87 dBW  −74.87 dBm From (19.9),   GTransponder dB  [EIRPsat ]dBW − [GT ]dBi − [PRsat ]dBW  65 dBW − 42 dBi + 104.87 dBW  127.87 dB ∼  128 dB Thus, the gain of the transponder is 128 dB. Block diagram of a typical satellite communication transponder with power lev- els at different stages based on the previous examples is shown in Fig. 19.10. The transponder consists of a preselect filter (PSF) at the input of the transponder to select the desired frequency band. The PSF restricts the noise and interferences outside the desired bandwidth entering into the receiver. The signals are received by the receiver, amplified by a low-noise amplifier (LNA) and then frequency translated by mixer and further amplified by IF amplifiers. A bandpass filter (BPF) is used just after the mixer to suppress the spurious products generated by the mixer. The output signals from the receiver are divided into several channels by channelization filters and then amplified to high power level using different transmitter chains. The channelized high power signals then combined in frequency domain by using an output mul- tiplexer. The multiplexed signal passes through a harmonic reject filter to provide sufficient rejection to the harmonics generated by the power amplifiers operated in their nonlinear region. Total gain of a transponder of about 130 dB is distributed among the receiver and transmitter. The gains of receive and transmit sections are about 50 and 84 dB, respectively, as shown in Fig. 19.10. The loss of the preselect filter and feeder cable in between antenna output and the receiver input needs to be taken into account to determine the overall system noise temperature. Similarly, loss of the output multiplexer, HRF and interconnecting cable up to antenna input are to be taken into account to determine final output power transmitted by the antenna to achieve required saturated EIRP. 598 19 Microwave Communication Systems Fig. 19.10 Block diagram of a satellite communication system with power levels Example 19.12 Calculate insertion loss and return losses of a 2-port network of S-parameter matrix: 0.05 − 80◦ 0.84 − 25◦ [S]  0.84 − 25◦ 0.05 − 80◦ Calculate insertion loss and return losses when such two networks connected in cascade. Solution Insertion loss (IL) of the network is ILdB  −20 × log(|S21 |)  −20 × log(0.84)  1.51 dB Return loss (RL) of the network is RLdB  −20 × log(|S11 |)  −20 × log(0.05)  26.02 dB ABCD parameters of the network can be calculated using (7.124) (1 + S11 )(1 − S22 ) + S12 S21 A 2S21 (1 + 0.05 − 80◦ )(1 − 0.05 − 80◦ ) + 0.84 − 25◦ × 0.84 − 25◦  2 × 0.84 − 25◦ 19.3 Satellite Communication Systems 599 (1 + 0.009 − 0.049i)(1 − 0.009 + 0.049i) + (0.761 − 0.355i) × (0.761 − 0.355i)  2 × (0.761 − 0.355i)  0.924 4.662◦ (1 + S11 )(1 + S22 ) − S12 S21 B  Zo 2S21 (1 + 0.05 − 80◦ )(1 + 0.05 − 80◦ ) − 0.84 − 25◦ × 0.84 − 25◦  50 × 2 × 0.84 − 25◦ (1 + 0.009 − 0.049i)(1 + 0.009 − 0.049i) + (0.761 − 0.355i) × (0.761 − 0.355i)  50 × 2 × (0.761 − 0.355i)  21.252 63.159◦ (1 + S11 )(1 − S22 ) − S12 S21 C  Yo 2S21 1 (1 + 0.05 − 80◦ )(1 − 0.05 − 80◦ ) − 0.84 − 25◦ × 0.84 − 25◦  × 50 2 × 0.84 − 25◦ 1 (1 + 0.009 − 0.049i)(1 − 0.009 + 0.049i) − (0.761 − 0.355i) × (0.761 − 0.355i)  × 50 2 × (0.761 − 0.355i)  0.010 75.463◦ (1 − S11 )(1 + S22 ) + S12 S21 D 2S21 (1 − 0.05 − 80◦ )(1 + 0.05 − 80◦ ) + 0.84 − 25◦ × 0.84 − 25◦  2 × 0.84 − 25◦ (1 − 0.009 + 0.049i)(1 + 0.009 − 0.049i) + (0.761 − 0.355i) × (0.761 − 0.355i)  2 × (0.761 − 0.355i)  0.924 4.662◦ Thus, ABCD parameters in matrix form of the 2-port network are 0.924 4.662◦ 21.252 63.159◦ [ABCD]  0.010 75.463◦ 0.924 4.662◦ [ABCD] parameters of such two 2-port networks connected in cascade will be given by [Ac Bc Cc Dc ]  [ABCD] × [ABCD] 0.924 4.662◦ 21.252 63.159◦ 0.924 4.662◦ 21.252 63.159◦  × 0.010 75.463◦ 0.924 4.662◦ 0.010 75.463◦ 0.924 4.662◦ 0.740 21.976◦ 39.283 67.821◦  0.018 80.125◦ 0.740 21.976◦ [S] parameters of the cascaded 2-port network can be derived using (7.125) Ac + Bc Yo − Cc Zo − Dc Cascaded S11  Ac + Bc Yo + Cc Zo + Dc 600 19 Microwave Communication Systems 0.740 21.976◦ + 0.786 67.821◦ − 0.910 80.125◦ − 0.740 21.976◦  0.740 21.976◦ + 0.786 67.821◦ + 0.910 80.125◦ + 0.740 21.976◦ 0.220 − 50.338◦  2.841 50.049◦  0.077 − 100.387◦ Thus, return loss (RL) of the cascaded network is Cascaded RL (in dB)  −20 × log(0.077)  22.22 dB 2 Cascaded S21  Ac + Bc Yo + Cc Zo + Dc 2  0.740 21.976◦ + 0.786 67.821◦ + 0.910 80.125◦ + 0.740 21.976◦ 2  2.841 50.049◦  0.704 − 50.049◦ Thus, insertion loss (IL) of the cascaded network is Cascaded IL (in dB)  −20 × log(0.704)dB  3.05 dB !! ( about (1.51 + 1.51) dB) This example shows that the insertion loss of the cascaded network (3.049 dB) is equal to sum of the insertion losses of the individual networks. Example 19.13 Calculate insertion loss and return losses of a 2-port network of S-parameter matrix: 0.30 − 90◦ 0.84 − 26◦ [S]  0.84 − 26◦ 0.30 − 90◦ Calculate insertion loss and return losses when such two networks connected in cascade. Solution Insertion loss (IL) of the network is IL (in dB)  −20 × log(|S21 |)  −20 × log(0.84)  1.51 dB 19.3 Satellite Communication Systems 601 Return loss (RL) of the network is RL (in dB)  −20 × log(|S11 |)  −20 × log(0.30)  10.46 dB ABCD parameters of the network can be calculated using (7.124) (1 + S11 )(1 − S22 ) + S12 S21 A 2S21 (1 + 0.3 − 90◦ )(1 − 0.3 − 90◦ ) + 0.84 − 26◦ × 0.84 − 26◦  2 × 0.84 − 26◦ (1 + 0 − 0.3i)(1 − 0 + 0.3i) + (0.755 − 0.368i) × (0.755 − 0.368i)  2 × (0.755 − 0.368i) ◦  0.966 5.961  (1 + S11 )(1 + S22 ) − S12 S21 B  Zo 2S21 (1 + 0.3 − 90◦ )(1 + 0.3 − 90◦ ) − 0.84 − 26◦ × 0.84 − 26◦  50 × 2 × 0.84 − 26◦ (1 + 0 − 0.3i)(1 + 0 − 0.3i) + (0.755 − 0.368i) × (0.755 − 0.368i)  50 × 2 × (0.755 − 0.368i) ◦  14.215 20.717  (1 + S11 )(1 − S22 ) − S12 S21 C  Yo 2S21 1 (1 + 0.3 − 90◦ )(1 − 0.3 − 90◦ ) − 0.84 − 26◦ × 0.84 − 26◦  × 50 2 × 0.84 − 26◦ 1 (1 + 0 − 0.3i)(1 − 0 + 0.3i) − (0.755 − 0.368i) × (0.755 − 0.368i)  × 50 2 × (0.755 − 0.368i) ◦  0.015 93.638  (1 − S11 )(1 + S22 ) + S12 S21 D 2S21 (1 − 0.3 − 90◦ )(1 + 0.3 − 90◦ ) + 0.84 − 26◦ × 0.84 − 26◦  2 × 0.84 − 26◦ (1 − 0 + 0.3i)(1 + 0 − 0.3i) + (0.755 − 0.368i) × (0.755 − 0.368i)  2 × (0.755 − 0.368i) ◦  0.966 5.961  602 19 Microwave Communication Systems Thus, ABCD parameters in matrix form of the 2-port network are 0.966 5.961◦ 14.215 20.717◦ [ABCD]  0.015 93.638◦ 0.966 5.961◦ [ABCD] parameters of such two 2-port networks connected in cascade will be given by [Ac Bc Cc Dc ]  [ABCD] × [ABCD] 0.966 5.961◦ 14.215 20.717◦  0.015 93.638◦ 0.966 5.961◦ 0.966 5.961◦ 14.215 20.717◦ × 0.015 93.638◦ 0.966 5.961◦ 0.911 25.027◦ 27.459 26.677◦  0.029 99.598◦ 0.911 25.027◦ [S] parameters of the cascaded 2-port network can be derived using (7.125) Ac + Bc Yo − Cc Zo − Dc Cascaded S11  Ac + Bc Yo + Cc Zo + Dc 0.911 25.027◦ + 0.549 26.677◦ − 1.437 99.598◦ − 0.911 25.027◦  0.911 25.027◦ + 0.549 26.677◦ + 1.437 99.598◦ − 0.911 25.027◦ 1.380 − 58.039◦  3.090 52.000◦  0.447 − 110.039◦ Thus, return loss (RL) of the cascaded network is Cascaded RLdB  −20 × log(0.447)  7.00 dB 2 Cascaded S21  Ac + Bc Yo + Cc Zo + Dc 2  0.911 25.027◦ + 0.549 26.677◦ + 1.437 99.598◦ − 0.911 25.027◦ 2  3.090 52.000◦  0.647 − 52.000◦ Thus, insertion loss (RL) of the cascaded network is Cascaded IL (in dB)  −20 × log(0.647)  3.78 dB !! (different from (1.51 + 1.51) dB) 19.3 Satellite Communication Systems 603 This example shows that the insertion loss of the cascaded network (3.78 dB) is more than the sum of the insertion losses of the individual networks. This is due to the poor port return losses (10.46 dB) of the individual networks. Example 19.14 Calculate insertion loss and return losses of the two 2-port networks P and Q of S-parameter matrixes: 0.6 − 91◦ 0.8 − 1◦ [S]P  0.8 − 1◦ 0.6 − 91◦ and 0.5 26◦ 0.9 − 63◦ [S]Q  0.9 − 63◦ 0.5 26◦ Calculate insertion loss and return losses when such two networks connected in cascade. Solution Insertion loss (IL) of the network P is IL (in dB)  −20 × log(|S21 |)  −20 × log(0.8)  1.938 dB Return loss (RL) of the network P is RL (in dB)  −20 × log(|S11 |)  −20 × log(0.6)  4.44 dB Insertion loss (IL) of the network Q is IL (in dB)  −20 × log(|S21 |)  −20 × log(0.9)  0.92 dB Return loss (RL) of the network Q is RL (in dB)  −20 × log(|S11 |)  −20 × log(0.5)  6.02 dB 604 19 Microwave Communication Systems ABCD parameters of the network P can be calculated using (7.124) (1 + S11 )(1 − S22 ) + S12 S21 A 2S21 (1 + 0.6 − 91◦ )(1 − 0.6 − 91◦ ) + 0.8 − 1◦ × 0.8 − 1◦  2 × 0.8 − 1◦ (1 − 0.01 − 0.6i)(1 + 0.01 + 0.6i) + (0.8 − 0.014i) × (0.8 − 0.014i)  2 × (0.8 − 0.014i)  1.250 0◦ (1 + S11 )(1 + S22 ) − S12 S21 B  Zo 2S21 (1 + 0.6 − 91◦ )(1 + 0.6 − 91◦ ) − 0.8 − 1◦ × 0.8 − 1◦  50 × 2 × 0.8 − 1◦ (1 − 0.01 − 0.6i)(1 − 0.01 − 0.6i) − (0.8 − 0.014i) × (0.8 − 0.014i)  50 × 2 × (0.8 − 0.014i)  36.409 − 90◦ (1 + S11 )(1 − S22 ) − S12 S21 C  Yo 2S21 1 (1 + 0.6 − 91◦ )(1 − 0.6 − 91◦ ) − 0.8 − 1◦ × 0.8 − 1◦  × 50 2 × 0.8 − 1◦ 1 (1 − 0.01 − 0.6i)(1 + 0.01 + 0.6i) − (0.8 − 0.014i) × (0.8 − 0.014i)  × 50 2 × (0.8 − 0.014i)  0.015 90◦ (1 − S11 )(1 + S22 ) + S12 S21 D 2S21 (1 − 0.6 − 91◦ )(1 + 0.6 − 91◦ ) + 0.8 − 1◦ × 0.8 − 1◦  2 × 0.8 − 1◦ (1 + 0.01 + 0.6i)(1 − 0.01 − 0.6i) + (0.8 − 0.014i) × (0.8 − 0.014i)  2 × (0.8 − 0.014i)  1.250 0◦ Thus, ABCD parameters in matrix form of the 2-port P network are 1.250 0◦ 36.409 − 90◦ [ABCD]P  0.015 90◦ 1.250 0◦ In similar way, ABCD parameters in matrix form of the 2-port Q network are 19.3 Satellite Communication Systems 605 0.516 − 3.535◦ 78.865 90.039◦ [ABCD]Q  0.009 92.513◦ 0.516 − 3.535◦ Therefore, [ABCD] parameter matrix of the cascaded networks P and Q can be written as [ABCD]PQ  [ABCD]P × [ABCD]Q 0.984 − 1.448◦ 79.817 90.880◦  0.020 90.060◦ 0.575 175.949◦ [S] parameters of the cascaded 2-port network can be derived using (7.125) A + BYo − CZo − D Cascaded S11  A + BYo + CZo + D 0.984 − 1.448◦ + 1.596 90.880◦ − 0.981 90.060◦ − 0.575 175.949◦  0.984 − 1.448◦ + 1.596 90.880◦ + 0.981 90.060◦ + 0.575 175.949◦ 1.658 22.379◦  2.540 81.295◦  0.653 − 58.916◦ Thus, return loss (RL) of the cascaded network is Cascaded RL (in dB)  −20 × log(0.653)  3.70 dB 2 Cascaded S21  A + BYo + CZo + D 2  0.984 − 1.448◦ + 1.596 90.880◦ + 0.981 90.060◦ + 0.575 175.949◦ 2  2.540 81.295◦  0.787 − 81.295◦ Thus, insertion loss (RL) of the cascaded network is Cascaded IL (in dB)  −20 × log(0.787)  2.08 dB ! ! (different from (1.938 + 0.92) dB) This example shows that the insertion loss (2.08 dB) of the cascaded network is less than the sum of the insertion losses of the individual networks. This is due to the poor port return losses of the individual networks. Examples 19.12, 19.13 and 19.14 show that in case of good return losses of the individual networks, insertion loss and gain of the combined (cascaded) network are the algebraic sum of the individual networks’ gain (loss), whereas, in case of poor return losses of the individual networks, combined (cascaded) gain (loss) may increase or decrease. 606 19 Microwave Communication Systems 19.4 Receiver Function of on-board receiver of a communication satellite is to amplify the received signal linearly with minimum possible addition of noise and translating the frequency band of the received signal to the required downlink frequency band using a local oscillator. Most important parameters of a communication receiver are operating frequency, bandwidth, gain, noise figure, linearity, spurious levels and frequency translation error [2–4]. Block diagram of a typical receiver with operating gain of about 50 dB and overall noise figure of 1.7 dB is shown in Fig. 19.11. Front end of the receiver consists of a low-noise amplifier (LNA) to achieve overall low-noise temperature of the system. The input matching network of an LNA corresponds to its optimum noise figure which is different from complex conjugate matching; thus, input VSWR of an LNA is poor. An isolator is used at input of the LNA to avoid mismatch of the receive antenna output to the receiver. A high-gain LNA (28 dB in this example) is used to minimize the noise contributions due to the losses of the following elements such as filters and mixer. Here, a three-stage low-noise amplifier using pHEMT device is used for total gain of 28 dB and noise figure of 1.7 dB. In general, double-balanced mixer is used to downconvert the receive frequency band to transmit IF frequency band using a local oscillator. The frequency of the local oscillator is the frequency difference in between receive and transmit frequencies. A bandpass filter (BPF) at the input of the mixer is used to pass only the required frequency band and rejecting the unwanted out-of-band frequencies including the rejection of image frequency band. The BPF at the output of the mixer is used to reject various mixing products generated by the mixer. An IF amplifier is used to provide rest of the gain required for the receiver. The IF amplifier is designed to extract maximum power gain from the device, which operates in its linear region with moderate noise figure. Thus, if stability criterion allows, the IF amplifiers are designed with complex conjugate matching at its input as well as output for achieving maximum power gain from the used device. The amplifiers may be realized in hybrid microwave integrated circuit (HMIC) or in monolithic microwave integrated circuit (MMIC). In case of HMIC implementation, discrete components including active devices are mounted on printed substrate to realize the complete circuit, whereas, in MMIC implementation, all passive and active components including matching elements are built in a single substrate. Due to the absence of packaging of individual components and interconnecting elements in MMIC realization, the frequency of operation and achievable bandwidth are more. Photograph of a 3-stage microwave amplifier, realized using discrete components, is shown in Fig. 19.12. Here, packaged pHEMT, chip resistors and chip capacitors are used as discrete components which are mounted in an alumina substrate on which matching transmission microstrip elements are printed. Photograph of a double-conversion receiver is shown in Fig. 19.13 which is realized in HMIC. 19.4 Receiver 607 Fig. 19.11 Block diagram of a single conversion communication receiver pHEMT Fig. 19.12 Photograph of a microwave amplifier realized in HMIC Example 19.3 Calculate overall noise figure (NF) and noise temperature of a receiver of block diagram as shown in Fig. 19.11. Consider loss of the isolator  0.2 dB. Solution Isolator loss: 0.2 dB LNA gain: 28 dB, LNA NF: 1.7 dB Filters and mixer combined loss: (1 + 8 + 1) dB  10 dB IF amplifier gain: 32 dB, IF amplifier NF: 3 dB Simplified block diagram of the receiver with gain and noise figures of individual modules is shown in Fig. 19.14. Using the following Friis formula for calculation of receivers overall noise figure (NFRX ), NF2 − 1 NF3 − 1 NF4 − 1 NFRX  NF1 + + + G1 G1 × G2 G1 × G2 × G3 0.479 9.0 0.995  1.047 + + + 0.955 0.955 × 631 0.955 × 631 × 0.1  1.5803 608 19 Microwave Communication Systems Fig. 19.13 Photograph of a communication receiver Fig. 19.14 Block diagram of the receiver for noise calculation [NFRX ]dB  1.987 dB Noise temperature TRx of the receiver is TRX  (NFRX − 1)To  (1.5803 − 1)290 K  168.3 K Example 19.4 Calculate overall third-order intermodulation product (IM3) of a receiver (shown in the block diagram in Fig. 19.11) at input power level of −76 dBm. The Po1dB (output power at 1-dB gain compression point) of LNA and IF amplifiers are 5 and +7 dBm, respectively. The PoIP3 (output third-order intercept point) of the mixer is 10 dBm. 19.4 Receiver 609 Solution Po1dB of LNA: +5 dBm Po1dB of IF amplifier: +7 dBm PoIP3 of mixer: +10 dBm Considering the I–O characteristic of LNA, mixer and IF amplifier can be expressed by a power series up to third order, the third-order intercept output power level PoIP3 can be written in terms of Po1dB as PoIP3  Po1dB + 10.63 Using this equation, PoIP3 of the LNA and IF amplifier are +15.63 and +17.63 dBm, respectively. Output power levels of each element at the input carrier power level of −76 dBm are shown in Fig. 19.15, considering all the elements are operating with constant gain. The third-order intermodulation power level PoIM 3 in dBc with respect to the output carrier level is given by (14.15b) PoIM 3  Po2f1 −f2 − Pof1  2 Pof1 − PoIP3 Here, Pof1  Pout . Therefore, the third-order intermodulation power level of LNA, mixer and IF amplifiers at the output of respective elements are PoIM 3 (LNA)  2(−48.2 − 15.6) dBc  −127.7 dBc ⇒ −175.9 dBm PoIM 3 (Mixer)  2(−57.2 − 10) dBc  −134.4 dBc ⇒ −191.6 dBm PoIM 3 (IF Amp)  2(−26.2 − 17.6) dBc  −87.7 dBc ⇒ −113.86 dBm Considering the linear gain of the IM 3 power levels by the following stages, the overall IM 3 level at the output of the receiver will be PoIM 3  (−175.9 + 22) dBm + (−191.6 + 31) dBm + (−113.86) dBm  −113.86 dBm Fig. 19.15 Block diagram of the receiver for IM3 calculation 610 19 Microwave Communication Systems The IM 3 level in dBc with respect to the carrier level at the output is (−113.86 + 26.2) dBc  −87.7 dBc. This shows that for this receiver configuration, the contribution of LNA and mixer on the overall IM 3 is negligible. It is fully governed by the nonlinearity of the final stage, i.e. IF amplifier. Therefore, if the receiver operates at higher input power levels say by 10 or 20 dB more than overall IM 3 level of the receiver will be more by 20 and 40 dB, respectively. Thus, for the receiver input power level of −66 and −56 dBm, the IM 3 level will be −67.7 and −47.7 dBc, respectively. 19.4.1 Local Oscillator The role of a local oscillator in satellite transponder is to provide a stable reference RF frequency with sufficient output power level to drive mixer circuit in its nonlinear region for frequency translation. The frequency of the local oscillator is the difference between the centre frequency of the uplink band and the centre frequency of the downlink frequency band. Frequency stability, spectral purity in terms of phase noise, and spurious products and output power level are the most important parameters of a local oscillator. Generally, in a satellite application, local oscillators of the desired frequency are derived from a reference low-frequency source. The stability of local oscillator is determined by the stability of the reference frequency source. Mostly, crystal-based oscillators are used as reference source for the local oscil- lators. Crystal cut in the form of a plate determines the fundamental frequency of oscillation. Frequency stability of a reference crystal oscillator is mainly influenced by the change of temperature and time. To minimize the influence of change of tem- perature, temperature-compensated crystal oscillator (TCXO) is used. Here, TCXO encases the oscillator circuit and temperature compensating networks in a closed container. Another option is the use of oven-controlled crystal oscillator (OCXO) as reference oscillator. Here, crystal oscillator and temperature-sensitive elements are kept in a thermally insulated container along with a heater. The heater maintained the inside temperature at oscillators’ minimum sensitive region. The stability of a local oscillator is specified as long-term stability over the lifetime and short-term stability over the specified operating temperature range. In general, for satellite transponders, temperature-controlled crystal oscillator (TCXO) of short-term stability of ±1 ppm (parts per million, i.e. error of 1 Hz in 1 MHz) and long-term stability of ±10 ppm for 15 years is used. In some applications, where more stable frequency is required, oven-controlled crystal oscillator (OCXO) of one order that has better stability com- pared to TCXO is used. Two types of local oscillators are realized. One is based on frequency multiplier and another is using phase-locked loop (PLL). 19.4 Receiver 611 (a) (b) Fig. 19.16 a Block diagram of a multiplier-based local oscillator. b Simple block diagram of a PLL-based local oscillator 19.4.1.1 Multiplier-Based Local Oscillator Block diagram of a multiplier-based LO is shown in Fig. 19.16a. Here, reference crystal oscillator frequency of 132 MHz and required LO output frequency of 3.3 GHz is considered. Two stages of X5 multiplication (overall X25) are done to generate 3.3 GHz frequency from the reference frequency of 130 MHz. Bandpass filters at different stages are used to suppress unwanted frequency components to a sufficient level (at least −60 dBc) with respect to the desired frequency component. Any unwanted frequency component of an LO acts as discrete frequency component and may severely affect the communication. Amplifiers (AMP1 and AMP2) are used before each multiplier circuit to provide required power level to the multiplying device to operate it highly nonlinear con- dition. The final amplifier (AMP3) is used to increase the power level of the final frequency component to achieve the desired output level required for operating mixer in the receiver in its nonlinear region. In general, the final amplifier stage operates in saturation region to provide nearly constant output power level over the operat- ing environmental (temperature, bias voltage variation, etc.) condition to ensure the operation of mixer in its fixed conversion gain/loss condition. 19.4.1.2 PLL-Based Local Oscillator Simplified block diagram of a phase-locked loop (PLL)-based local oscillator is shown in Fig. 19.16b. It consists of a phase detector, loop filter, frequency divider and voltage-controlled oscillator (VCO). Generally, varactor diode is used to achieve voltage-dependent frequency of the VCO. Output frequency of the VCO is divided by a frequency divider and fed back to one input of the phase detector circuits. The phase detector compares the output frequency from the divider to the reference frequency of the reference oscillator. A phase error signal is generated by the phase detector 612 19 Microwave Communication Systems and creates a signal whose magnitude is proportional to the phase error. This phase error signal is then low-pass filtered by the loop filter and fed to the control input of the VCO. The control signal controls the output frequency of the VCO. At the locked condition of the PLL, the two inputs to the phase detector are in-phase and the output frequency is equal to the reference oscillator frequency multiplied by the divider ratio, N. 19.5 Satellite Transmitter The transmitter section amplified the channelized signal to the required RF power level before transmitting through a transmit antenna. As shown in Fig. 19.10, this section consists of a driver amplifier (DA) to boost the signal to proper drive level required for high power amplifier (HPA) which ultimately provides the required transmit power level. In a communication system, major contributions of nonlin- earities are due to the final power amplifier that affects the overall communication performance severely [5–7]. A linearizer also used at input of the HPA to minimize the effect of nonlinearity of the HPA on transmitted signal. The driver amplifier (DA) is also called channel amplifier (CAMP) for its function to amplify the channelize frequency band. The CAMPs consist of several control systems for on-board con- trolling the transponder gain by ground command or automatically by sensing its RF power level. The channel amplifier and linearizer are low-power systems, and thus these are realized very compactly using solid-state technology. However, depending on down- link frequency and required output power level, two types of high power amplifiers are used: Solid-state power amplifier (SSPA) and travelling wave tube amplifier (TWTA). Use of TWTAs is the only option in case of requirement of higher RF over high microwave frequency range. However, due to the advancement of solid- state device technology, it can provide required RF power level at least over lower microwave frequency range. Thus, SSPAs are preferable for its compact size and better linearity compared to TWTAs. 19.5.1 Driver Amplifier (DA) Function of an on-board driver amplifier (DA) for a communication satellite is to amplify the channelized signal linearly, i.e. without much distortion, for providing sufficient drive level to the high power amplifier [8]. It has also provision for on-board gain setting of the transponder by ground command or automatically by sensing its RF power level. A driver amplifier for satellite transponder has two operating modes: fixed gain mode (FGM) and automatic level control (ALC) mode. Most important parameters of a driver amplifier are its operating frequency, bandwidth, gain, linearity, output power, gain setting range in FGM and dynamic range in ALC mode. In a 19.5 Satellite Transmitter 613 Fig. 19.17 Block diagram of a satellite channel amplifier satellite transponder, a driver amplifier amplifies the signal after channelization, and thus it is also called channel amplifier (CAMP). Block diagram of a channel amplifier is shown in Fig. 19.17. The CAMP consists of RF circuits and bias and control circuits. The RF lineup consists of several amplifier stages (A1–A5) to achieve required gain. Digital attenuators (DAT1 and DAT2) are used for commandable gain setting of the CAMP by issuing digital command. The analog attenuators (AAT1 and AAT2) are used to provide attenuation automatically by detecting the power level using RF power detector (DET). The detected voltage at the output of the detector amplified by a differential DC amplifier and applied to the control terminal of the analog attenuators through a commandable analog switch. In ALC mode of operation, the output of the DC amplifier will be connected to the analog attenuators. Another input of the differential DC amplifier is connected to a temperature-controlled voltage to keep the output RF power level constant in ALC mode over the operating temperature range. In FGM operation, the control terminals of the analog attenuators will be connected to temperature-controlled voltage to keep the constant gain over the operating temperature range. The digital control circuit processes the command data and generates appropriate control data to select the mode of operation (in between ALC and FGM) and to provide commandable gain setting in both the operation modes. An amplitude tilt active equalizer (EQ) is used to achieve broadband frequency response. The amplifiers and both types of attenuators are distributed in the lineup to achieve required noise figure and linearity of the CAMP over its entire gain (dynamic range) setting conditions for both the operation modes. Photograph of a Ku-band CAMP of about 60 dB gain is shown in Fig. 19.18. The amplifier and attenuator modules are realized using MMIC technology. The full RF 614 19 Microwave Communication Systems Fig. 19.18 Photograph of a Ku-band CAMP Fig. 19.19 Frequency response of the CAMP with and without equalizer circuit is packaged in three compartments with narrow slits (0.7 mm × 2 mm) through which the circuits are interconnected using gold ribbons. The slit acts as waveguide of cut-off frequency below 10 GHz, and thus provides high isolations for the operating frequency in between two adjacent compartments, which prevents waveguide mode of propagation in operating frequency band, and thus ensures overall stability of the CAMP. Frequency response of the CAMP with and without equalizer is shown in Fig. 19.19. It shows that the use of equalizer improves the gain flatness from 6 to 1.5 dB over the frequency range of 10.5–13 GHz. In fixed gain mode (FGM) operation, the digital attenuators DAT1 and DAT2 are used to control the gain of the CAMP to set the saturation flux density (SFD) of the transponder and also to oper- ate the transponder at required power back-off condition. The typical gain setting is about 30-dB in steps of 1-dB. I–O characteristics of the channel amplifier in FGM operation for different gain setting conditions are shown in Fig. 19.20. The output of the CAMP saturated under higher input power level is due to the saturation of the final amplifier stage. Under nominal operating condition, the channel amplifier always operates in its linear I–O characteristic region. 19.5 Satellite Transmitter 615 Fig. 19.20 I-O characteristic of CAMP in FGM operation for different gain setting condition Fig. 19.21 I-O characteristic of CAMP in ALC mode operation for different attenuation in DAT2 In ALC mode of operation, the output power will remain constant irrespective of its input power level with the specified ALC dynamic range. Typical ALC dynamic range is about 30 dB. Here, the RF power level is detected before the digital attenuator DAT2 to control the output power level in ALC mode for operating the HPA back-off condition if required. Final amplifier stage is used after the detector and DAT2 to enable the use of low-power device for the final amplifier stage meeting the output power requirement. I–O characteristics of the CAMP in its ALC mode of operation are shown in Figs. 19.21 and 19.22 for setting of the digital attenuators DAT2 and DAT-1, respectively. In ALC mode of operation, the digital attenuator DAT2 is used to provide adjustable (variable) constant output power level as shown in Fig. 19.21. Typically, 15 dB attenuation range in steps of 0.5 dB is kept for this purpose. The digital attenuator DAT1 is used to slide the ALC range keeping the same ALC dynamic range as shown in Fig. 19.22. This provides the flexibility of SFD range setting of the transponder in ALC mode. 616 19 Microwave Communication Systems Fig. 19.22 I-O characteristic of CAMP in ALC mode operation for different attenuation in DAT1 19.5.2 Travelling Wave Tube Amplifier (TWTA) Travelling wave tube amplifier (TWTA) is one of the most economically costliest subsystems which used most critical but highly matured technology for realization [9]. In most of the satellite transponders, TWTAs are used as high power amplifier (HPA) for its ability to provide higher output power with higher DC-to-RF efficiency at higher frequency of operation over broader frequency band compared to solid- state power amplifiers (SSPAs). A TWTA consists of travelling wave tube (TWT) and electronic power conditioner (EPC). The EPC provides required DC voltage and currents to the TWT and provides various controls and protection mechanisms of the TWT. In a satellite transponder, TWTAs amplify signal taking from the output of a channel amplifier and provides required output power. In most of the cases, a predistortion linearizer is used in between the channel amplifier and TWTA to minimize the effect of nonlinearity of a TWTA on communication system. In travelling wave tube amplifier, amplification of microwave signal takes place due to the continued interaction between the wave and the high-energy electron beam travelling along the signal. Functional diagram of a TWT is shown in Fig. 19.23. Structurally, a TWT can be divided into three sections: electron gun, slow-wave structure and collector. The electron is generated by heating a cathode which travel towards anode due to high electric field generated by applying a very high potential difference in between anode and cathode. The electron beam after passing through the helix is collected at the collector. To fulfil the requirement of continued interac- tion of waves with the electron beam for long time, the microwave signal is passed through a slow-wave structure through which the electron beam flows and the elec- tron beam is focused applying a longitudinal static magnetic field using permanent magnets as shown in Fig. 19.23. The helical slow-wave structure slowed down the microwave signal with its phase velocity of about cp/2π r, where p is the pitch of the helix of radius r. Due to the propagation of electromagnetic wave along the helix, a longitudinal electric field will be generated. This time-varying electric field results in velocity modulation in the electron beam passing through the helix. This velocity modulation will result in bunching of electrons in regular intervals of one wavelength 19.5 Satellite Transmitter 617 Fig. 19.23 Functional diagram of TWT Fig. 19.24 Typical I–O, gain and phase characteristic of TWTA of the applied signal. Thus, with the condition of equal phase velocity of microwave signal and electron beam, a continuous interaction takes place between the beams and the waves in the helix and bunches grow as the beam moves ahead. This continued interaction results in the amplification of microwave signal flowing through the helix by picking power from the electron beam. The amplified microwave signal is then coupled out of the helix at the output port. The collected electrons at the collector dissipate its rest of the energy at the collector. Most important RF performance parameters of a TWTA are operating frequency, bandwidth, saturated output power, DC-to-RF efficiency and linearity. Capability to provide higher output power of about 250 W at S-band and 150 W at Ka-band with DC-to-RF efficiency about 65–70% for space grade TWTAs makes them very attractive for using as HPA in high-power satellite transponders. However, it has comparatively poor linearity, size and more mass compared to SSPAs. I–O char- acteristic, gain and phase dependency on RF power level are shown in Fig. 19.24. Typical values of gain compression and total phase shift are about 6.6 dB and 45°, respectively. These nonlinearities lead to distortion in amplified signal. One way to minimize the effect of these nonlinearities on amplified signal is to operate the TWTA at power back-off condition. However, efficiency decreases with the increase of back- off of TWTA. Practically, a linearizer is used with TWTA to minimize the effect of nonlinearity of TWTAs and thus minimizes the requirement of TWTA back-off. 618 19 Microwave Communication Systems 19.5.3 Solid-State Power Amplifier (SSPA) Solid-state power amplifiers (SSPAs) are used as high power amplifiers where output power requirement is less. SSPAs are advantageous for its better linearity, lower mass and smaller footprint area compared to TWTAs. Wherever SSPAs are capable to provide the required RF power level, it is preferable to use SSPAs. For example, at L-, S- and C-bands, SSPAs as well as TWTAs are used depending on the output power requirement. However, beyond C-band there is no option other than using TWTAs for space communication due to the non-availability of SSPAs with required power level. With the present technology, space qualified SSPA of output power up to 60 W is realized using GaAs-based MESFET/HFET power devices. With the advancement of GaN technology, presently devices are available to realize space qualified SSPAs of output power level about 250 W at L- and S-bands and 120 W at C-band. Block diagram of a solid-state power amplifier of output RF power 53 dBm (200 W) is shown in Fig. 19.25. It consists of RF section, electronic power condi- tioner (EPC) and bias and temperature control circuits. The RF section consists of medium power amplifier and high power amplifier stages to provide output power level of 200 W with overall gain of 53 dB. Important parameters of an SSPA are operating frequency, bandwidth, output power level, DC-to-RF efficiency and lin- earity. To achieve higher efficiency, in general, the final two amplifier stages operate in Class-AB mode with harmonic tuning. However, the medium power amplifier stages operate in Class-A mode to achieve overall better linearity without sacrificing overall efficiency. The efficiency of a final stage amplifier at its maximum allowable operation power level is comparable to the efficiency of a TWTA which is about 70%. However, overall efficiency of an SSPA is significantly reduced due to the use of other amplifying stages to achieve gain comparable with a TWTA. For example, the efficiency of a C-band 40 W GaAs-based SSPA (including EPC) is about 45%. Another important aspect of design of a power amplifier is the protection of power devices under intentional or unintentional RF overdrive condition. Lineup of the power amplifier is so selected that full range of overdrive cannot pass to final ampli- fier stage. Another option is the use of open-loop limiter or closed-loop overdrive protection circuit to protect the devices from overdrive condition [10]. The bias and control circuits provide required bias voltages/currents to all the devices which are temperature-controlled to achieve temperature-compensated gain as well as output power level. Practically, it is required to increase the drain voltage of the final amplifying FET device with the increase of temperature to keep the output power constant over the operating temperature range. Photograph of an SSPA realized in hybrid microwave integrated circuit (HMIC) is shown in Fig. 19.26. Near the saturation point, SSPA behaves differently from TWTA. Typical I–O, gain and phase characteristics of an SSPA are shown in Fig. 19.27. In case of TWTA, the output power decreases when it operated beyond its saturated point; however, for an SSPA, it remains nearly constant. SSPAs are not permitted to operate beyond its 2-dB gain compression point to ensure its reliability for space use. Excessive gate current may flow when a power FET operates beyond its 2-dB gain compression 19.5 Satellite Transmitter 619 Fig. 19.25 Block diagram of a solid-state HPA Fig. 19.26 Photograph of a solid-state power amplifier point. Excessive flow of gate current for a power FET over a prolonged duration may lead to failure of the device. However, low-power FETs may be operated at more gain compression without compromising its life. Typical total phase shift for an SSPA is about 20° up to its 2-dB gain compression point. Another important point is to be noted that in case of SSPAs the phase of the output signal increases with the increase of power level as shown in Fig. 19.27, whereas it decreases in case of TWTAs as seen in Fig. 19.24. Thus, it is clear that SSPAs are in general more linear than TWTAs when both are operated at their respective rated output power conditions. In practice, an SSPA 620 19 Microwave Communication Systems Fig. 19.27 Typical I–O characteristics of an SSPA without linearizer has about equally good overall linearity as linearity of a linearized TWTA. Thus, there is very less scope for improvement of linearity of SSPAs using a linearizer. More precisely, using linearizer, there is large scope of phase as well amplitude nonlinearity improvement for TWTAs. However, there is only scope of phase nonlinearity improvement for an SSPA. 19.6 Linearizer Among the various types of linearizers, predistortion (PD) linearizers are mostly used for satellite communication due to their simplicity, low power consumption, and ability to linearize over wide bandwidth of a power amplifier operating at near saturation condition [11–15]. Predistortion linearizer creates inverse nonlinearity of the transmitting amplifiers such as of TWTAs or SSPAs in order to compensate for the distortion. It is able to function as standalone unit and can be cascaded in between CAMP and HPA with proper input and output power level matching. Block diagram of a broadband predistortion linearizer is shown in Fig. 19.28. Heart of the linearizer is the distortion generator module. In general, it is realized using Schottky and p-i-n diodes in a vector modulator configuration as discussed in Chap. 16 to generate required amplitude and phase nonlinearities. Two amplitude tilt active equalizers (EQ1 and EQ2) are used to make the linearizer broadband. An analog attenuator (AAT1) is used to compensate overall gain variation of the equalizer over its operating temperature range. The amplifier modules (A1, A2 and A3) are used to match the linearizer’s input and output power levels with the output power level of CAMP and input power level of HPA in addition to compensate loss of the distortion generator, equalizers and attenuator. Photograph of a linearizer operating in Ku-band is shown in Fig. 19.29. The linearizer is realized using MMIC amplifier modules and other circuits realized on alumina substrate using HMIC technology. The distortion generator unit is realized as vector modulator with one arm nonlinear circuit using Schottky barrier diodes and the other arm with a linear circuit using p-i-n diodes. The 19.6 Linearizer 621 Fig. 19.28 Block diagram of a broadband predistortion linearizer Fig. 19.29 Photograph of a Ku-band linearizer amplitude tilt active equalizers and analog attenuator are realized using p-i-n diodes as voltage/current variable resistors. Temperature compensation of the unit is done using optimum load-line bias technique of Schottky barrier and p-i-n diodes as well as providing temperature-controlled voltage/current to the analog attenuator (AAT1). Typical nonlinear amplitude and phase performances of a Ku-band linearizer cas- caded with a TWTA are shown in Fig. 19.30. It shows that gain compression of a non- linearized TWTA improves from 6.5 to ±0.7 dB when it is cascaded and optimized with linearizer. Similarly, nonlinearized TWTA’s total phase shift improves from 45° to ±4°. Thus, amplitude and phase nonlinearity of linearized TWTA (LTWTA) improves significantly and comparable, even better than linearity of a nonlinearized SSPA. Two-tone third-order intermodulation levels (IM 3 ) of TWTA and LTWTA are shown in Fig. 19.31. It shows that there is no significant improvement of IM 3 level near saturation region of the TWTA. However, there is a significant IM 3 improve- ment at around 7 dB input back-off (IBO) of each carrier which corresponds to 4 dB IBO of the TWTA with respect to its saturation point. 622 19 Microwave Communication Systems Fig. 19.30 Typical I–O, gain and phase characteristic of TWTA and Linearized TWTA Fig. 19.31 Typical 3rd order IM levels of TWTA and LTWTA 19.7 Microwave Power Module (MPM) Microwave power module (MPM) is the technology where both solid-state and vac- uum tube technologies are combined to realize very compact with low mass transmit- ter module to provide highest possible output power with highest efficiency [16, 17]. The solid-state technology is advantageous for its miniaturized configuration using MMIC technology with excellent linearity. However, its output power giving capa- bility and efficiency is less compared to TWTA. Presently, space qualified MMICs with medium output power up to 10 W are available. On the other hand, TWTAs are capable of providing highest output power level with high DC-to-RF efficiency. But its linearity is poor and also size is more in case of high-gain TWTAs. In MPM, short- length TWT is used as final power amplifier with lower gain which needs input power about 1–10 W which is achievable from solid-state MMIC configuration. Thus, rest 19.7 Microwave Power Module (MPM) 623 Fig. 19.32 Block diagram of microwave power module (MPM) of the transmitter gain with medium-level output power is achieved using solid-state technology in MMICs. To improve the linearity of the total system, a predistortion linearizer is realized in solid-state technology cascaded at suitable position as shown in Fig. 19.32. To minimize the overall mass and footprint area, a single electronic power conditioner is used for all these microwave subsystem and a single mechani- cal assembly including all the RF units and the EPC. To realize a microwave power module, challenge is the thermal management due to the very compact assembly of the unit. EMI/EMC-related issues are also critical due to the proximity of the small signal and high power units in a single package. 19.8 Multiport Amplifier (MPA) Multiport amplifiers (MPAs), also known as matrix amplifiers, have multiple inputs and multiple outputs with several high power amplifiers connected in parallel. These are used to provide flexibility in terms of power allocation, since combined output power from parallel-connected several amplifiers is shared between the output ports. Here, RF power can be allocated among the output ports as per requirements of a communication system. Therefore, the combined power of all the high power amplifiers is available for any output port provided that the other ports do not require any power at the same time. The multiport amplifier configurations are very useful in case of multibeam satellite communication systems, where beam-to-beam service requirements such as number of users/data rate changes dynamically over time. In case of multiport amplifier configuration, all the power amplifiers provide same RF output power irrespective of different powers required by the different output ports. Thus, operating conditions (input/output back-off) of all the amplifiers remain the same though RF powers taken from different output ports are different. 624 19 Microwave Communication Systems Fig. 19.33 Block diagram of a 4 × 4 MPA with signal flow paths (bold lines) when power fed to only Port-I 1 A multiport amplifier (MPA) consists of an array of power amplifiers (PAs) in parallel and a pair of complementary Butler matrix networks that consist of 90o hybrid networks [18]. Schematic diagram of a 4 × 4 MPA is shown in Fig. 19.33. The 4 × 4 MPA consists of four high power amplifiers connected in parallel using 4 × 4 input and 4 × 4 output networks. The signal at each input in the MPA is divided into four signals (in general n signals) with particular phase relationships. These signals are amplified separately in each power amplifier and are recombined in the output network. Thus, signal at each input is amplified by all the power amplifiers, however, assembled at the corresponding output ports. Figure 19.33 shows the flow of signal when it is applied to the input port-1. Thus, the signal is applied in port-1 amplified by all the amplifiers and then recombined at the output port-1. In a similar way, applied signal at input ports 2, 3 and 4 is amplified by all the amplifiers and recombined at the respective output ports 2, 3 and 4. In ideal condition, signal applied to the port-1 becomes available at the output port-1 only. Thus, no part of the signal applied at the input port-1 will be available at the output ports 2, 3 and 4. To achieve this ideal performance from an MPA, it is important to equalize the amplitude and to synchronize the transmitted phase among the signals. Practically, infinite isolation is not possible due to various nonideal electrical characteristics of each signal path. The finite isolation leads to signal loss as well as interference among signals coming from different input ports. Another drawback is the unwanted multicarrier operation of the power amplifiers. This multicarrier operation reached as all the input signals are amplified at each power amplifier even when a single carrier is fed at each input. The multiport amplifier is adapted to applications that require flexibility in terms of power allocation at its different output ports and is really advantageous if the input signals are already multicarrier. Design of compact input and output Butler matrix networks with minimum inser- tion loss and proper amplitude and phase matching is crucial. The input network operates at lower RF power level, and thus loss of the input network can be easily 19.8 Multiport Amplifier (MPA) 625 (a) (b) Fig. 19.34 4 × 4 Butler matrix using 3-dB 90° hybrid couplers a with crossover, b planner structure without crossover Fig. 19.35 8 × 8 Butler matrix using 3-dB 90° hybrid couplers compensated by providing more gain at the driver amplifier stages. It is always prefer- able to realize input network in planar transmission line configuration very compactly with the compromise of insertion loss. Figure 19.34 shows planner configuration of a 4 × 4 Butler matrix with and without crossover connection and Fig. 19.35 shows planner configuration of an 8 × 8 Butler matrix suitable for realization in microstrip line configuration. Any loss of the output butler network decreases total available output power. Generally, waveguide-based output network is designed to achieve low insertion 626 19 Microwave Communication Systems Fig. 19.36 Block diagram of the system of Example 19.5 loss, though it is bulky. Followings are the various examples related to microwave communication systems and subsystems. Example 19.5 Derive the overall noise figure of a communication system which consists of a receiver and transmitter as shown in Fig. 19.36. Calculate the overall system noise figure for gain and noise figure of the receiver: 50 and 2 dB, respectively, and gain and noise figure of the transmitter: 80 and 20 dB, respectively. Solution Gain of the receiver  GRX  50 dB  105 Noise figure of the receiver  NFRX  2 dB  1.585 Gain of the transmitter  GTX  80 dB  108 Noise figure of the transmitter  NFTX  20 dB  100 Suppose the available noise power at the input of the receiver is Pni ; this is the thermal noise over the noise bandwidth B of the receiver and is given by Pni  kTo B Thus, the noise power output, PnoRX , of the receiver can be written as PnoRX  PniTX  (Pni ) × (GRX NFRX )  Pni × GRX + (NFRX − 1)Pni × GRX The first term is the output noise due to the amplification of the input available thermal noise, and the second part is the noise added by the receiver (Fig. 19.36). In the similar way, the total noise power at the output of the transmitter can be written as PnoTX  PniTX × GTX + (NFTX − 1)Pni × GTX PnoTX  Pni × GRX × NFRX × GTX + (NFTX − 1)Pni × GTX (19.10a) Suppose NFRXTX is the overall noise figure of the system. The overall gain of the system is GRX × GTX . Thus, the noise power at the output of the transmitter can be written as PnoTX  (Pni ) × (GRX × GTX × NFRXTX ) (19.10b) 19.8 Multiport Amplifier (MPA) 627 Fig. 19.37 Block diagram of the system of Example 19.6 Comparing (19.10a) and (19.10b), the overall system noise figure can be written as (NFTX − 1) NFRXTX  NFRX + (19.11) GRX This is known the Friis’s formula, already derived in Chap. 14. This formula shows that in case of sufficiently high receiver gain, the overall noise factor is dominated by the noise factor of the receiver. Putting the values, the overall noise factor of the system (100 − 1) NFRXTX  1.585 +  1.586 100,000  2.003 dB This shows that though the noise figure of the transmitter is poor, its effect on the overall system is negligible due to the high gain of the receiver. Example 19.6 Derive overall power-added efficiency of a communication system which consists of a receiver and transmitter as shown in Fig. 19.37. Calculate the overall power-added efficiency of the system for gain- and power-added efficiency of the receiver: 50 dB and 1%, respectively, and gain- and power-added efficiency of the transmitter: 80 dB and 50%, respectively. Solution Gain of the receiver  GRX  50 dB  105 Power-added efficiency of the receiver  ηRX  1%  0.01 Gain of the transmitter  T RX  80 dB  108 Power-added efficiency of the transmitter  ηTX  50%  0.5 Suppose DC power to the receiver  PDCRX DC power to the transmitter  PDCTX Input RF power of the receiver  PINRX Input RF power of the transmitter  PINTX  PORX Output RF power of the transmitter  POTX 628 19 Microwave Communication Systems From the definition of power-added efficiency, PORX − PINRX ηRX  (19.12a) PDCRX POTX − PINTX ηTX  (19.12b) PDCTX The overall power-added efficiency of the system ηRXTX can be written as POTX − PINRX ηRXTX  (19.12c) PDCRX + PDCTX Putting GRX  PORX /PINRX , GTX  POTX /PINTX and using (19.12a), (19.12b), the overall efficiency of the system can be written as 1 GRX − 1 1 (GTX − 1)GRX 1  × + × (19.13) ηRXTX GRX GTX − 1 ηRX GRX GTX − 1 ηTX Putting GRX  105 , ηRX  0.01, TRX  108 and ηTX  0.5 1 105 − 1 1 108 − 1 105 1  × + × ηRXTX 1013 − 1 0.01 1013 − 1 0.5 or ηRXTX  0.5  50% This shows that though the power-added efficiency of the receiver is poor, its effect on the overall system is negligible due to the high gain of the transmitter and receiver. Considering GRX  1 and GTX  1, the overall power-added efficiency of the system (19.13) can be written as 1 GRX 1 GTX GRX 1  × + × (19.14a) ηRXTX GRX GTX ηRX GRX GTX ηTX 1 1 1 1  × + (19.14b) ηRXTX GTX ηRX ηTX or ηRXTX  ηTX (19.14c) Example 19.7 Derive worst-case overall two-tone third-order intermodulation level of a communication system which consists of a receiver and transmitter as shown in Fig. 19.38. Calculate the overall two-tone third-order intermodulation level of the system for IM 3 of the receiver and transmitter, which are (a) 20 and 10 dBc (b) 10 and 10 dBc, respectively. 19.8 Multiport Amplifier (MPA) 629 Fig. 19.38 Block diagram of the system of Example 19.7 Solution Suppose Output RF power of the receiver  PORX  PINTX Output RF power of the transmitter  POTX Therefore, level of IM3RX at the output of the receiver is PORX × 10−(IM3RX /10) (19.15a) Here, consider that the third-order intermodulation signal generated by the receiver will be linearly amplified by the transmitter. Thus, at the output of the transmitter, the third-order intermodulation signal level will be 10−(IM3RX /10) × PORX × GTX + POTX × 10−(IM3TX /10) (19.15b) −(IM3RX /10) −(IM3TX /10)  POTX × 10 + 10 (19.15c) Here, the first term is the contribution by the receiver and the second term is the contribution by the transmitter. Thus, the overall IM 3 level POTX × 10−(IM3RX /10) + 10−(IM3TX /10) IM3RXTX  POTX  10−(IM3RX /10) + 10−(IM3TX /10) [IM3RXTX ]dB  10 log 10−(IM3RX /10) + 10−(IM3TX /10) (19.15d) (a) Putting the value of IM3RX  20 dBc and IM3TX  10 dBc [IM3RXTX ]dB  10 log 10−(20/10) + 10−(10/10)  10 log(0.01 + 0.1)  10 log(0.01 + 0.1)  9.59 dB (b) Putting the value of IM3RX  10 dBc and IM3TX  10 dBc [IM3RXTX ]dB  10 log 10−(10/10) + 10−(10/10) 630 19 Microwave Communication Systems  10 log(0.1 + 0.1)  −6.99 dB This example shows that effect of nonlinearity of receivers and transmitters equally affects the overall linearity of the system. Example 19.8 Derive expressions for output powers of a transmitter of I-O charac- teristic governed by vo  a1 vi + a3 vi3 for a two-tone carrier input, vi  A cos ω1 t + B cos ω2 t. Considering the transmitter is matched at its input and output ports with Ro  50 , and a1  10, a3  −0.04, calculate the output power levels corre- sponding to the fundamental and third-order harmonics for the following cases: (a) For total input power level of 0 dBW of equal power levels of two carriers. (b) For total input power level of 0 dBW of unequal power levels of two carriers by 6 dB. (c) For total input power level of 0 dBW of unequal power levels of two carriers by 10 dB. Also, plot gains of both the carriers over the total input power level of −20 to 0 dBW for the case-b and case-c. Solution I–O characteristic of the transmitter is given by vo  a1 vi + a3 vi3 (19.16) The two carriers’ input excitation is vi  A cos ω1 t + B cos ω2 t (19.17) Therefore, the output is given by vo  a1 (A cos ω1 t + B cos ω2 t) + a3 (A cos ω1 t + B cos ω2 t)3 (19.18) 3 3 vo  a3 A2 B + a3 AB2 2 2  3 3 + a1 A + a3 A3 + a3 AB2 cos ω1 t 4 2   3 3 + a1 B + a3 B3 + a3 A2 B cos ω2 t 4 2     1 1 + a3 A3 cos 3ω1 t + a3 B3 cos 3ω2 t 4 4   3 + a3 A2 B [cos(2ω1 − ω2 )t + cos(2ω1 + ω2 )t] 4   3 + a3 AB2 [cos(2ω2 − ω1 )t + cos(2ω2 + ω1 )t] (19.19) 4 19.8 Multiport Amplifier (MPA) 631 Amplitude of the output voltage corresponding (vo1 ) to carrier ω1 is given by 3 3 voω1  a1 A + a3 A3 + a3 AB2 4 2 Thus, the output power corresponding to carrier ω1 is given by    2 1 voω1 2 3 3 Poω1  √  a1 A + a3 A + a3 AB /2Ro 3 2 (19.20a) Ro 2 4 2 Similarly, the output power corresponding to carrier ω2 is given by    2 1 voω2 2 3 3 Poω2  √  a1 B + a3 B3 + a3 A2 B /2Ro (19.20b) Ro 2 4 2 The output power corresponding to each third-order intermodulation product (2ω1 ± ω2 ) is given by  2 3 Po(2ω1 ±ω2 )  a3 A2 B /2Ro (19.21a) 4 Similarly, the output power corresponding to each third-order intermodulation product (2ω1 ± ω2 ) is given by  2 3 Po(2ω2 ±ω1 )  a3 AB2 /2Ro (19.21b) 4 The input power corresponding to carrier ω1 is given by √ 2 Piω1  A/ 2 /Ro  A2 /2Ro (19.22a) The input power corresponding to carrier ω2 is given by √ 2 Piω2  B/ 2 /Ro  B2 /2Ro (19.22b) Therefore, total input power is given by PiTotal  A2 + B2 /2Ro (19.23) (a) For equal power levels of both the carriers with PiTotal  0 dBW  1 W. Thus, from (19.23), A  B  7.071 V 632 19 Microwave Communication Systems Fig. 19.39 Power levels under two-tone excitation of equal input power levels of Example 19.8 Here, A  B  7.071 V and a1  10, a3  −0.04 and Ro  50 . From (19.22a) and (19.22b), Piω1  Piω2  A2 /2Ro  7.0712 /100  0.5 W  −3.01 dBW From (19.20a) and (19.20b),  2 3 3 Poω1  Poω2  a1 A + a3 A3 + a3 AB2 /2Ro 4 2  2 3 3  10 × 7.071 − × 0.04 × 7.0713 − × 0.04 × 7.071 × 7.0712 /100 4 2  15.13 W  11.80 dBW From (19.21a), (19.21b),  2 3 Po(2ω1 ±ω2 )  Po(2ω2 ±ω1 )  a3 A2 B /2Ro 4  2 3  × 0.04 × 7.0713 /100 4  1.12 W  051 dBW The input and output power levels of the two equal power carriers are shown in Fig. 19.39. Under the two-tone carrier excitation of equal power levels, the output power levels for both the fundamental frequency components are same and also the power levels of both the third-order IMD components are same. The levels of the third-order IMD components are 19.8 Multiport Amplifier (MPA) 633 Po(2ω1 ±ω2 ) − Poω1  Po(2ω2 ±ω1 ) − Poω2  (0.51) dBW − 11.80 dBW  −11.29 dBc (b) For unequal power levels by 6 dB with PiTotal  0 dBW  1 W. From (19.22a) and (19.22b), Piω1 − Piω2  10 log A2 /2Ro − 10 log B2 /2Ro  6 dB (19.24a) And from (19.23), PiTotal  10 log A2 + B2 /2Ro  0 dBW (19.24b) From (19.24a) and (19.24b), A  8.94 V, B  4.48 V The input powers corresponding to carriers ω1 and ω2 are given by, Piω1  A2 /2Ro  8.942 /100  0.799 W  −0.973 dBW Piω2  B2 /2Ro  4.482 /100  0.201 W  −6.973 dBW From (19.20a),  2 3 3 Poω1  a1 A + a3 A3 + a3 AB2 /2Ro 4 2  2 3 3  10 × 8.94 − × 0.048.94 − × 0.04 × 8.94 × 4.48 /100 3 2 4 2  32.71 W  15.15 dBW From (19.20b),  2 3 3 Poω2  a1 B + a3 B3 + a3 A2 B /2Ro 4 2  2 3 3  10 × 4.48 − × 0.04 × 4.48 − × 0.04 × 8.94 × 4.48 /100 3 2 4 2  4.25 W  6.29 dBW From (19.21a),  2 3 Po(2ω1 ±ω2 )  a3 A2 B /2Ro 4 634 19 Microwave Communication Systems Fig. 19.40 Power levels under two-tone excitation of unequal input levels by 6 dB of Example 19.8  2 3  × 0.04 × 8.942 × 4.48 /100 4  1.15 W  0.62 dBW From (19.21b),  2 3 Po(2ω2 ±ω1 )  a3 AB2 /2Ro 4  2 3  × 0.04 × 8.94 × 4.482 /100 4  0.29 W  −5.38 dBW The input and output power levels of the two unequal power level by 6 dB are shown in Fig. 19.40. The difference of output power levels among the fundamental carriers increases to 8.86 dB from the difference of 6 dB at input. This is due to the power transfer to the various harmonics and intermodulation components. This phenomenon is known as ‘power robbing’ when multiple carriers amplified by an amplifier operating in its nonlinear (gain compression) region. Though both the carri- ers pass through the same amplifier, they experience different levels of amplification. Gain responses of both the carriers are shown in Fig. 19.41 over the input power level of −20 to 0 dBW. At lower power levels where the amplifier operates in linear region, the gains of both the carriers are same. But over the nonlinear region of the 19.8 Multiport Amplifier (MPA) 635 Fig. 19.41 Gain of the carriers with unequal input power levels by 6 dB of Example 19.8 amplifier, the weaker signal amplifies less compared to the stronger signal as shown in Fig. 19.41. (c) For unequal power levels by 10 dB with PiTotal  0 dBW  1 W. From (19.22a) and (19.22b), Piω1 − Piω2  10 log A2 /2Ro − 10 log B2 /2Ro  10 dB (19.24a) And from (19.23), PiTotal  10 log A2 + B2 /2Ro  0 dBW (19.24b) From (19.24a) and (19.24b), A  9.535 V, B  3.015 V The input powers corresponding to carriers ω1 and ω2 are given by Piω1  A2 /2Ro  9.5352 /100  0.909 W  −0.414 dBW Piω2  B2 /2Ro  3.0152 /100  0.091 W  −10.414 dBW From (19.20a),  2 3 3 Poω1  a1 A + a3 A + a3 AB /2Ro 3 2 4 2  2 3 3  10 × 9.535 − × 0.04 × 9.5353 − × 0.04 × 9.535 × 3.0152 /100 4 2  41.142 W  16.14 dBW 636 19 Microwave Communication Systems From (19.20b),  2 3 3 Poω2  a1 B + a3 B + a3 A B /2Ro 3 2 4 2  2 3 3  10 × 3.015 − × 0.04 × 3.015 − × 0.04 × 9.535 × 3.015 /100 3 2 4 2  1.660 W  2.20 dBW From (19.21a),  2 3 Po(2ω1 ±ω2 )  a3 A2 B /2Ro 4  2 3  × 0.04 × 9.5352 × 3.015 /100 4  0.68 W  −1.70 dBW From (19.21b),  2 3 Po(2ω2 ±ω1 )  a3 AB2 /2Ro 4  2 3  × 0.04 × 9.535 × 3.0152 /100 4  0.07 W  −11.70 dBW The input and output power levels of the two carriers with unequal power levels by 10 dB are shown in Fig. 19.42. Here, it can be noted that due to the power robbing phenomenon, the difference of output power levels of the fundamental carrier increases to 13.94 dB compared to the difference of 10 dB at input. Gain responses of both the carriers are shown in Fig. 19.43 over the input power level of −20 to 0 dBW. It can be noted that the gain of the weaker carrier becomes 3.94 dB lower compared to the gain of the stronger carrier at 0 dBW total input power level. Thus, weak signal becomes further weak when it passes through a nonlinear (gain compression region) amplifier in the presence of stronger signal. Example 19.9 Write the expressions for output powers of a transmitter of IO characteristic governed by vo  a1 vi + a3 vi3 for a two-tone carrier input, vi  A cos ω1 t + B cos ω2 t. Considering the transmitter is matched at its input and out- put ports with Ro  50 , and a1  4, a3  0.02, calculate the output power levels corresponding to fundamental components and third-order harmonics for the following cases: (a) For total input power level of 0 dBW of equal power levels of two carriers. (b) For total input power level of 0 dBW of unequal power levels of two carriers by 10 dB. 19.8 Multiport Amplifier (MPA) 637 Fig. 19.42 Power levels under two-tone excitation of unequal input levels by 10 dB of Example 19.8 Fig. 19.43 Gain of the carriers with unequal input levels by 10 dB of Example 19.8 Also, plot gains of both the carriers over the total input power level of −20 to 0 dBW for the case-b. Solution I–O characteristic of the transmitter is given by vo  a1 vi + a3 vi3 Under the two-carrier input excitation of vi  A cos ω1 t + B cos ω2 t, the output power corresponding to carrier ω1 is given by (19.20a) 638 19 Microwave Communication Systems    2 1 voω1 2 3 3 Poω1  √  a1 A + a3 A + a3 AB /2Ro 3 2 Ro 2 4 2 Similarly, the output power corresponding to carrier ω2 is given by (19.20b)    2 1 voω2 2 3 3 Poω2  √  a1 B + a3 B3 + a3 A2 B /2Ro Ro 2 4 2 The output power corresponding to each third-order intermodulation product (2ω1 ± ω2 ) is given by (19.21a)  2 3 Po(2ω1 ±ω2 )  a3 A2 B /2Ro 4 The output power corresponding to each third-order intermodulation product (2ω1 ± ω2 ) is given by (19.21b)  2 3 Po(2ω2 ±ω1 )  a3 AB2 /2Ro 4 The input power corresponding to carrier ω1 is given by (19.22a) √ 2 Piω1  A/ 2 /Ro  A2 /2Ro The input power corresponding to carrier ω2 is given by (19.22b) √ 2 Piω2  B/ 2 /Ro  B2 /2Ro Therefore, total input power is given by (19.23) PiTotal  A2 + B2 /2Ro (a) For equal power levels of both the carriers with PiTotal  0 dBW  1 W. From (19.23), A  B  7.071 V Here, A  B  7.071 V and a1  4, a3  +0.02 and Ro  50 . From (19.22a) and (19.22b), Piω1  Piω2  A2 /2Ro  7.0712 /100  0.5 W  −3.01 dBW 19.8 Multiport Amplifier (MPA) 639 Fig. 19.44 Power levels under two-tone excitation of equal input levels of Example 19.9 From (19.20a) and (19.20b),  2 3 3 Poω1  Poω2  a1 A + a3 A3 + a3 AB2 /2Ro 4 2  2 3 3  4 × 7.071 + × 0.02 × 7.0713 + × 0.02 × 7.071 × 7.0712 /100 4 2  19.53 W  12.91 dBW From (19.21a) and (19.21b),  2 3 Po(2ω1 ±ω2 )  Po(2ω2 ±ω1 )  a3 A2 B /2Ro 4  2 3  × 0.02 × 7.0713 /100 4  0.28 W  −5.51 dBW The input and output power levels of the two equal power carriers are shown in Fig. 19.44. Under the two-tone carrier excitation of equal power levels, the output power levels for both the fundamental frequency components are same and also the power levels of both the third-order IMD components are same. The levels of the third-order IMD components are Po(2ω1 ±ω2 ) − Poω1  Po(2ω2 ±ω1 ) − Poω2  −5.51 dBW − 12.91 dBW  −18.42 dBc 640 19 Microwave Communication Systems (b) For unequal power levels by 10 dB with PiTotal  0 dBW  1 W. Thus, from (19.22a) and (19.22b), Piω1 − Piω2  10 log A2 /2Ro − 10 log B2 /2Ro  10 dB And from (19.23), PiTotal  10 log A2 + B2 /2Ro  0 dBW From (19.24a) and (19.24b), A  9.535 V, B  3.015 V The input powers corresponding to carriers ω1 and ω2 are given by Piω1  A2 /2Ro  9.5352 /100  0.909 W  −0.414 dBW Piω2  B2 /2Ro  3.0152 /100  0.091 W  −10.414 dBW From (19.20a),  2 3 3 Poω1  a1 A + a3 A3 + a3 AB2 /2Ro 4 2  2 3 3  4 × 9.535 + × 0.02 × 9.535 + × 0.02 × 9.535 × 3.015 /100 3 2 4 2  28.88 W  14.61 dBW From (19.20b),  2 3 3 Poω2  a1 B + a3 B + a3 A B /2Ro 3 2 4 2  2 3 3  4 × 3.015 + × 0.02 × 3.0153 + × 0.02 × 9.5352 × 3.015 /100 4 2  4.28 W  6.32 dBW From (19.21a),  2 3 Po(2ω1 ±ω2 )  a3 A2 B /2Ro 4  2 3  × 2.02 × 9.5352 × 3.015 /100 4  0.17 W  −7.72 dBW 19.8 Multiport Amplifier (MPA) 641 Fig. 19.45 Power levels under two-tone excitation of unequal input levels by 10 dB of Example 19.9 From (19.21b),  2 3 Po(2ω2 ±ω1 )  a3 AB2 /2Ro 4  2 3  × 0.02 × 9.535 × 3.0152 /100 4  0.02 W  −17.72 dBW The input and output power levels of the two unequal power levels by 10 dB are shown in Fig. 19.45. Here, the difference of output power levels of the fundamental carrier deceases to 8.29 dB from the difference of 10 dB at input. Gain responses of both the carriers are shown in Fig. 19.46 over the input power level of −20 to 0 dBW. It can be noted that the gain of the weaker carrier becomes 1.71 dB higher compared to the gain of the stronger carrier at 0 dBW total input power level. Example 19.10 A transceiver consists of a receiver, ALC driver amplifier and high power amplifier as shown in Fig. 19.47. The ALC driver amplifier controls the gain of the transponder from 110 to 140 dB depending on its input power level to operate the transponder at saturation condition over its 30 dB dynamic range. Calculate the noise and carrier powers at the transponder output when the transponder operates in saturated condition for (a) minimum and (b) maximum gain considering the following parameters: System noise temperature (Ts ) 650 K Transponder noise bandwidth (BN ) 100 MHz Saturated gain of the HPA 50 dB Saturated output power of HPA 140 W 642 19 Microwave Communication Systems Fig. 19.46 Gain of the carriers with unequal input levels by 10 dB of Example 19.9 Fig. 19.47 Gain of the carriers with unequal input levels by 6 dB Plot carrier and noise power at the output of the transponder over its gain of 110–140 dB considering the transponder are operating at saturation condition. Solution The system noise temperature (Ts ): 650 K Input noise power (PN in ) over the frequency band of 100 MHz: PN in  kTs BN  1.38 × 10−23 × 650 × 100 × 106 W  −90.47 dBm (19.25) (a) At minimum gain of the transponder: Total input noise power PN in  −90.47 dBm Gain of the transponder: 110 dB Thus, Total output noise power PN out  (−90.47 + 110) dBm  19.53 dBm  0.09 W 19.8 Multiport Amplifier (MPA) 643 Saturated output power of the transponder:  140 W Considering saturated output power of the HPA consists of amplified noise and carrier power only (neglecting intermodulation powers), the carrier power can be written as  output saturated power − output noise power  140 W − 0.09 W  139.91 W  51.46 dBm (b) At maximum gain of the transponder: Total input noise power PN in  −90.47 dBm Gain of the transponder: 140 dB Thus, Total output noise power PN out  (−90.47 + 140) dBm  49.53 dBm  89.70 W Saturated output power of the transponder:  140 W Considering saturated output power of the HPA consists of amplified noise and carrier power only (neglecting contribution of intermodulation powers) and assuming saturated power of HPA is same for single and multicarrier conditions, the carrier power can be written as  output saturated power − output noise power  140 W − 89.70 W  50.30 W  47.02 dBm Plot of the output carrier and noise power of the transponder over its entire gain range of 110–140 dB is shown in Fig. 19.48. Example 19.11 Show the use of 3-dB 90° hybrid coupler to separate RHCP and LHCP signals from an antenna. Solution Electric field of a right-hand circular polarized (RHCP) signal can be expressed as E(r, t)  Ea sin(ωa t − ka z)êx + Ea cos(ωa t − ka z)êy (19.26) 644 19 Microwave Communication Systems Fig. 19.48 Carrier and noise powers versus gain at output of the transponder Fig. 19.49 Scheme for the separation of LHCP and RHCP signals using 3-dB 90° hybrid And the electric field of another left-hand circular polarized (LHCP) signal can be expressed as E(r, t)  −Eb sin(ωb t − kb z)êx + Eb cos(ωb t − kb z)êy (19.27) Two input ports of a 3-dB 90° hybrid coupler are connected to the two ports of an antenna as shown in Fig. 19.49. Here, horizontal and vertical feeds are connected to the two inputs (1 and 2) of the hybrid. At the port-1 (say, z  0), the signal voltage can be written as v1 (t)  va sin(ωa t) − vb sin(ωb t) (19.28) Similarly, at the port-2 (say, z  0), the signal voltage can be written as v2 (t)  va cos(ωa t) + vb cos(ωb t) (19.29) 19.8 Multiport Amplifier (MPA) 645 Thus, the signal voltage at port-3 can be written as va vb v3 (t)  √ sin(ωa t − π) − √ sin(ωb t − π ) 2 2 va π  vb π + √ cos ωa t − + √ cos ωb t − 2 2 2 2 va vb va vb  − √ sin(ωa t) + √ sin(ωb t) + √ sin(ωa t) + √ sin(ωb t) 2 2 2 2 vb vb  √ sin(ωb t) + √ sin(ωb t) 2 2 √  2vb sin(ωb t) (19.30) This shows that the output port-3 of the hybrid contains only the frequency com- ponent corresponding to LHCP. Thus, the signal voltage at the port-4 can be written as va π vb π v3 (t)  √ sin ωa t − − √ sin ωb t − 2 2 2 2 va vb + √ cos(ωa t − π ) + √ cos(ωb t − π ) 2 2 va vb va vb  − √ cos(ωa t) + √ cos(ωb t) − √ cos(ωa t) − √ cos(ωb t) 2 2 2 2 va va  − √ cos(ωa t) − √ cos(ωa t) 2 2 √  2va sin(ωa t) (19.31) This shows that the output port-4 of the hybrid contains only the frequency com- ponent corresponding to RHCP. In a similar way, it can be shown that two signals of different frequency bands can be transmitted one in LHCP and other in RHCP through a single antenna using a 3-dB 90° hybrid. References 1. Wang C-X et al (2014) Cellular architecture and key technologies for 5G wireless communi- cation networks. IEEE Commun Mag, Feb 2014 2. Pandian JD, Baker L, Cortes G, Goldsmith PF, Deshpande AA, Ganesan R, Hagen J, Locke L, Wadefalk N, Weinreb S (2006) Low-noise 6-8 GHz receiver. IEEE Microw Mag 7:74–84 3. Garg VK, Singh RV, Jain VK, Bera SC (1999) New amplifier design eliminates chip capacitors. IETE Tech Rev 16(6):197–201 4. Bera SC, Shah LB, Raval DU, Pandey S, Kumar V, Singh S, Das DK (2012) Design and development of V&W band amplifier modules. In: Proceedings of international conference on microwaves, antenna, propagation & remote sensing, 2012, Jodhpur, India 646 19 Microwave Communication Systems 5. Buch SD, Bera SC (2013) Transponder nonlinearity characterization & mitigation techniques: present scenario & future trends. In: International workshop on sensor network and wireless communication, ADIT, Oct 2013 6. Yadav SP, Bera SC (2014) Nonlinearity effect of high power amplifiers in communication sys- tems. In: International conference on advances in communication and computing technologies (ICACACT), Aug 2014 7. Yadav SP, Bera SC (2014) Nonlinearity effects of power amplifiers in wireless communication systems. In: Proceedings of IEEE international conference on electronics, communication and computational engineering (ICECCE 2014), Hosur, India, pp 1011–1016, Nov 2014 8. Bera SC, Singh RV, Garg VK (2006) Design and temperature compensation of a Ku-band channel amplifier with ALC for a satellite transponder. Microw J 49(4):68–82 9. Hu Y, Feng J (2016) The development and new trends of microwave vacuum electronic devices. In: IEEE international conference on emerging technologies (ICET), pp 1–5, Oct 2016 10. Bera SC, Singh RV (2004) A temperature-compensated closed loop overdrive level controller for microwave solid-state power amplifiers. Microw J 47(4):114–122 11. Yamauchi K, Mori K, Nakayama M, Mitsui Y, Takagi T (1997) A microwave miniaturized linearizer using a parallel diode with a bias feed resistance. IEEE Trans Microw Theory Tech 45(12):2431–2435 12. Bera SC, Bhardhwaj PS, Singh RV, Garg VK (2003) A diode linearizer for microwave power amplifiers. Microw J 46(11):102–113 13. Bera SC, Singh RV, Garg VK (2004) A compact Ku-band linearizer for space application. In: Proceedings of Asia Pacific microwave conference, 2004, pp 37–38 14. Bera SC, Singh RV, Garg VK (2008) Diode-based predistortion lineariser for power amplifiers. IEE Electron Lett 44(2):125–126 15. Bera SC, Kumar V, Singh S, Das DK (2013) Temperature behavior and compensation of diode- based predistortion linearizer. IEEE Microw Wirel Compon Lett 23(4):211–213 16. Abrams RH, Parker RK (1993) Introduction to the MPM: what it is and where it might fit. IEEE MTT-S Int Microw Symp Dig 1:107–110 17. Kowalczyk R, Zubyk A, Meadows C et al (2016) High efficiency E-band MPM for communi- cations application. In: 17th IEEE international vacuum electronics conference, pp 513–514, 2016 18. Bera SC, Singh RV, Garg VK (2008) Modified Wilkinson power divider with harmonic sup- pression characteristic. Microw J, Nov 2008