GSM Transceiver Design Optimization

Abstract

_________________________________________________________________________________ ABSTRACT—Transceiver design is a tradeoff between several parameters that impact on the transceiver specifications. These parameters are also interrelated such that an increase in one parameter impacts on the possible values of the second parameter and these determine the specifications achievable. This work presents the design of a GSM transceiver system and presents that achieved results in comparison to the ETSI GSM specifications. The parameters were all subjected to the same percentage variation and Matlab software was used to develop algorithms for determining the variations and interrelationships between the parameters .One key criteria in transceiver design is the Noise figure and the results show that the frontend comprising of the Duplexer, LNA, the Mixer has the greatest impact on the transceiver design while the power amplifier noise figures have very minimal effect of the overall system noise figure. It also shows the power amplifier as the unit that consumes the most power. From the results, several power saving techniques such as sleep modes targeted at the Power amplifiers can be implemented without having any significant effect on the transceiver speed, The results also show an inverse relationship between the overall system gain and the dynamic range of the transceiver such that to achieve a higher dynamic range, the system gain would have to be reduced.. This work enables an understanding of the relationships between the different parameters of the GSM transceiver design

Full text

Asian Journal of Applied Sciences (ISSN: 2321 – 0893)
Volume 04 – Issue 04, August 2016
Asian Online Journals (www.ajouronline.com) 955
GSM Transceiver Design Optimization
Francis E. Idachaba1.* and O. O. Oni2
1 Department of Electrical and Information Engineering
Covenant University
Ota, Nigeria
2 Department of Electrical and Information Engineering
Covenant University
Ota, Nigeria
*Corresponding author’s email: idachabafe [AT] yahoo.com
_________________________________________________________________________________
ABSTRACT—Transceiver design is a tradeoff between several parameters that impact on the transceiver
specifications. These parameters are also interrelated such that an increase in one parameter impacts on the possible
values of the second parameter and these determine the specifications achievable. This work presents the design of a
GSM transceiver system and presents that achieved results in comparison to the ETSI GSM specifications. The
parameters were all subjected to the same percentage variation and Matlab software was used to develop algorithms
for determining the variations and interrelationships between the parameters .One key criteria in transceiver design is
the Noise figure and the results show that the frontend comprising of the Duplexer, LNA, the Mixer has the greatest
impact on the transceiver design while the power amplifier noise figures have very minimal effect of the overall system
noise figure. It also shows the power amplifier as the unit that consumes the most power. From the results, several
power saving techniques such as sleep modes targeted at the Power amplifiers can be implemented without having any
significant effect on the transceiver speed, The results also show an inverse relationship between the overall system
gain and the dynamic range of the transceiver such that to achieve a higher dynamic range, the system gain would
have to be reduced. . This work enables an understanding of the relationships between the different parameters of the
GSM transceiver design
Keywords—GSM, transceiver, Noise figure, Gain
_________________________________________________________________________________
1. INTRODUCTION
Radio receiver design requires a compromise between the various specifications of the system. The parameters
involved are interrelated together both in direct and indirect proportions. This linkages and interdependence coupled with
the availability of different types of the same components with different specifications, makes the design of radio
receivers and transceivers a process of compromise between the different specifications.
The low noise amplifier is desired to have a very high gain so as to reduce the effect of the noise figure of the filters
and mixer stages. A high gain however affects the mixers linearity limiting its performance [1] The system gain is
desired to be high so that the transceivers can cover longer distances but the dynamic range of the receiver is reduced by
a high system gain. This entire interrelationship between the different blocks of the transceiver makes it necessary to
perform optimization on the system design to determine the parameter changes required so as to achieve system
specifications as close as possible to the target specification.
1.1 Transceivers
Transceivers can be described as a system comprising of a direct coupled receiver and transmitter systems. Figs 1(a)
and (b) is a combination of a receiver and a transmitter to produce a single down/up conversion transceiver based in the
super heterodyne topology. Figure 1(a) is a transceiver used for the uplink direction from the GSM mobile to the base
station or a satellite uplink while 1 (b) is for transmission from the satellite down to the mobile. This is known as
Downlink. Both configurations can be combined in a single transceiver block to create a bidirectional transceiver system

Asian Journal of Applied Sciences (ISSN: 2321 – 0893)
Volume 04 – Issue 04, August 2016
Asian Online Journals (www.ajouronline.com) 956
IRFilter IF Filter IF Amp Filter PA
VCO
LNA
Figure 1(a): Transceiver block diagram (Uplink)
IRFilterIF FilterIF Amp
Filter
PA
VCO
LNA
Figure 1(b): Transceiver block diagram (Downlink)
1.2 Noise figure
Noise is a vital factor in digital communication system as it gives rise to bit error due to the fact that the information
which is being transmitted and received may be incorrect or lost [2].In addition Signal to Noise Ratio (SNR) at the output
of receiving system is another very significant criterion. It is a function of the transmit output power, Gain of transmit
and receive antennas, receiver noise and pathloss [3].The overall noise figure of the cascaded system can be calculated
using the Friis equation and the noise figure and gain values of each block in the system.
G1, F1 G2. F2
Input Output
Gn. Fn
Figure 2: Noise Figure, F for cascaded network
In general, cascade of n devices is given by friss equation
n
n
GGG F
GG
F
G
F
..........2121
3
1
2
1n-1 1
..........
11
F F





Where
F = Noise figure of the Receiver
S I = Signal power at the input
G = Gain of the receiver
1.3 Sensitivity
Sensitivity is the ability of the receiver to reliably detect the minimum signals in a system by specifying the strength

Asian Journal of Applied Sciences (ISSN: 2321 – 0893)
Volume 04 – Issue 04, August 2016
Asian Online Journals (www.ajouronline.com) 957
of the minimum signal at the input. Minimum Detectable Signal (MDS) determines the sensitivity of the receiver and is
given as [4]:
)( min
SNRBTFKMDS SB

BdBSNRdBFdBmMDS 10log10)()(174)( min 
Where
SNR min =Minimum signal to noise ratio
F= Noise figure
Ts =Absolute temperature of the receivers input (⁰ K)
B= Receiver Bandwidth
Ks = Boltzmann’s constant 1.38*10-23Joules/K
2. TRANSCEIVER DESIGN PROCESS
The design of the transceiver involves selection of the various application specific integrated circuits (ASICs) such
that the overall system specification achieved will be within acceptable limits of the target specifications. The target
specifications for the GSM standard are as described by the ETSI and shown in Table 1
Table 1: ETSI 05.05 specifications
Parameters
Specification(DCS in brackets)
Sensitivity
-102 dBm (-100dBm)
Maximum receive signal strength
-15dBm
Noise figure
9.98dB(11.8dB)
C/N for BER performance
9dB
IIP3
-19.5 dBm
P1dB
-29.5dBm
Dynamic range
87dB
Up link
1710-1785MHz
Downlink
1805- 1880MHz
Channel band width
200KHz
2.1 Chip Selection
The various ASIC chips for the implementation of each of the blocks of the system to meet the ETSI specification are
identified and used in the system design.

Asian Journal of Applied Sciences (ISSN: 2321 – 0893)
Volume 04 – Issue 04, August 2016
Asian Online Journals (www.ajouronline.com) 958
Table 2: Selected Chips for the Transceiver UPLINK (1710-1785MHz)
Block
CHIP
G (dB)
N.F (dB)
Freq Range
OIP3
LNA
HMC375LP3
17.5
0.9
1.7-2.2GHz
34
SAW Image
reject filter
SAFCC1G74KA0
T00
-4.2
4.5
1710 – 1785MHz
100
Down
conversion
Mixer
HMC380QS16G
11
9
RF=1.7 - 2.2GHz
IF= 50 - 300MHz
19 (IIP3)
IF Filter
855625
-4.2
4.2
190MHz
(B/W 200KHz)
100
Gain Block
ADL5530
16
2.5
0 – 1GHz
(B/W 1GHz)
37dBm
Up conversion
Mixer
MAX 2039
-7.1
7.3
RF = 1.7 -2.2GHz
LO = 1.5 – 2.0GHz
IF = 0 – 350MHz
33.5dBm(IIP3)
SAW Image
reject filter
SAFCC1G74KA0
T00
-4.2
4.5
1710 – 1785MHz
100
The Power
Amplifier
HMC457Q16G
27dB
6dB
1.7 – 2.2GHz
46dBm (IIP3)
DOWNLINK (1805 – 1880MHz)
From the frequency distribution of the system, the same IF frequency value is used for the uplink and downlink, the
chips that differentiate both links are the image reject filters.
2.2 The SAW IR
The chip selected for Implementing the Image reject frequency for the uplink is the SAWEP1G84CQ0F00
2.3 The Voltage Controlled Oscillator (VCO)
With the choice of 190MHz (0.19GHz) for IF, the VCO is required to have the following frequency range.
RF – LO = IF
LOmin = RFmin – IF
LOmax = RFmax – IF
Uplink
LOmin = 1.710GHz – 0.19GHz = 1.520GHz
LOmax = 1.785GHz – 0.19GHz = 1.595GHz
Downlink
LOmin = 1.805GHz – 0.19GHz = 1.615GHz
LOmax = 1.880GHz – 0.19GHz = 1.690GHz
From the chip specifications for the VCO, the required LO frequencies for an IF of 190MHz is from 1.52GHz to a
maximum of 1.69GHz. A chip that will give this frequency value at low voltage is desired. The T0M9211 is the optimum
choice.

Asian Journal of Applied Sciences (ISSN: 2321 – 0893)
Volume 04 – Issue 04, August 2016
Asian Online Journals (www.ajouronline.com) 959
2.4 The Duplexer
The transceiver will utilize the same antenna for both receive and transmit so a duplexer would be required to
separate both channels. The duplexer chosen for the system is the ADF1800.
The transceiver designed based on these chips achieved the following specifications.
Table 3: Design specifications achieved and target specifications
Parameters
Design values (Achieved)
Target Specification
(DCS in brackets)
Sensitivity
-107.93 dBm
-102dBm(-100dBm)
Maximum receive signal
strength
-69.93 dBm
-15dBm
Noise figure
4.07dB
9.98dB(11.8dB)
C/N for BER performance
9dB
9dB
IIP3
-59.88dBm
-19.5 dBm
P1dB
-69.88dBm
-29.5dBm
Dynamic range
38dB
87dB
Uplink
1710-1785MHz
1710-1785MHz
Downlink
1805-1880MHz
1805- 1880MHz
Channel band width
200KHz
200KHz
From the results obtained, the transceiver designed achieved a better noise figure value than the target specification
but the IIP3, P1dB and dynamic range were below the target specification values.
3. DESIGN OPTIMIZATION USING MATLAB
Optimization in transceiver design involves a means of identifying the optimum value of the various components. The
optimization is done to determine the specification of components that can be used to achieve or approach the target
specification of the GSM receiver specifications [5].
There are three basic approaches used for circuit optimization and these are simulation based approach, equation
based approach and geometric programming [6,7,8,9]. The equation based approach is utilized using the Matlab software
to generate new specification based on the variation of the parameter values of each block.
The optimization procedure is as listed below.
(1) Determine the system specifications
(2) Apply the same percentage variation to a particular specification (gain, noise factor) of each block sequentially.
(3) Determine which component and which specification had the greatest impact on the transceiver’s overall
specifications.
(4) Based on steps 2 and 3 determine the component and parameter to alter and the amount of alteration required to
achieve the desired transceiver results.
From the optimization procedure listed above a variation value of 20% was selected to show the trend of the effect of
each component parameter on the overall system specifications. Using the Matlab software, the noise factor and the gain
values of each component were varied and their results are plotted to show their effect on the system specifications.
4. RESULTS AND DISCUSSION
The following results were obtained when the noise figure of each block was varied by a 20% change in value.
Plotting the graph to determine which component noise figure had the greatest effect on the overall system noise figure
yields the following results.

Asian Journal of Applied Sciences (ISSN: 2321 – 0893)
Volume 04 – Issue 04, August 2016
Asian Online Journals (www.ajouronline.com) 960
Variation of Noise factor
3.9
4
4.1
4.2
4.3
4.4
4.5
4.6
4.7
0 2 4 6 8 10 12 14
Block number
Noise Figure(dB)
Series1
DR(dB)
40.35
40.4
40.45
40.5
40.55
40.6
40.65
40.7
40.75
40.8
40.85
0 2 4 6 8 10 12 14
DR(dB)
From the graph, the first three blocks are the most critical as they have the greatest effect on the system noise
figure with a percentage variation of up to 15.38% for the LNA.
The effect of the variation of the gain of each component on both the dynamic range and the noise figure
yields the following results:
Effect of v ariation of the Block Gain parammee rs on
the system NF(dB)
3.6
3.65
3.7
3.75
3.8
3.85
3.9
3.95
4
4.05
G0 G1 G2 G3 G4 G5 G6 G7 G8 G9 G10 G11 G12
Blocks
NF(dB)
NF(dB)
The graph also shows that the gains of the first three blocks of the transceiver are the most critical and an increase in
the gain value leads to a reduction in the noise figure. From the graphs, an increase in the gain of blocks 4 to 12 had no
impact on the system noise figure.
From the formula, the dynamic range is given by:
Dynamic range = (2/3)*(11P3 – MDS)
4.1 Dynamic range optimization.
From the equation of the dynamic range, it is evident that the transceiver gain is related to the system dynamic range.
Variation of the system gain from 40dB to 100dB shows the following variation of the system gain from 40dB to 100Db

Asian Journal of Applied Sciences (ISSN: 2321 – 0893)
Volume 04 – Issue 04, August 2016
Asian Online Journals (www.ajouronline.com) 961
40 50 60 70 80 90 100
30
35
40
45
50
55
60
65
70
75 sysGain Vs DRange
sysGain in dB
Dynamic Range in dB
Figure 3: system gain Vs Dynamic range
From graph shown above in figure 3, the desired dynamic range and the corresponding system gain can be
determined. The result from the graph shows that a reduction in the gain leads to an increase in the dynamic range while
the results in table 1 shows that the gain of components after the IF stage have little effect on the system noise figure,
P1dB and sensitivity.
Eliminating one of the Power amplifiers and rerunning the program yielded the following
Gain
NF
MDS
Sensitivity
Dynamic
range
63.66
4.03
-113.96
-107.96
57.42
The results show an increase of the dynamic range to 57.42dB with the gain reduced to 63.66dB, the system NF
remained unaffected by the change in system gain. Table 4 shows the parameters achieved after the optimization
Table 4: Parameters achieved after optimization of the Transceiver
Parameters
Optimized values achieved
Target Specification (DCS in brackets)
Sensitivity
-107.96 dBm
-102 dBm (-100dBm)
Maximum receive signal strength
-50.54 dBm
-15dBm
Noise figure
4.03dB
9.98dB(11.8dB)
C/N for BER performance
9dB
9dB
IIP3
-27.82dBm
-19.5 dBm
P1dB
-37.82dBm
-29.5dBm
Dynamic range
57.42dB
87dB
Uplink
1710-1785MHz
1710-1785MHz
Downlink
1805-1880MHz
1805- 1880MHz
Channel band width
200KHz
200KHz
5. CONCLUSION
From the results in Table 4. the values of the components (gain and noise figure) have significant effects on
the overall system parameters. Components placed after the mixers have little effect on the system parameters.
The result also shows that the dynamic range is inversely related to the system gain. The dynamic gain can thus
be increased by reducing the power amplifier gain value. This reduction in gain can be made up for by the use of
highly directional antennas.

Asian Journal of Applied Sciences (ISSN: 2321 – 0893)
Volume 04 – Issue 04, August 2016
Asian Online Journals (www.ajouronline.com) 962
6. REFERENCES
[1] Parul Sharma1, Mrs J.Manjula, “design of low power and low noise figure gilbert mixer”, International Journal of
Advanced Research in Electrical, Electronics and Instrumentation Engineering vol. 2, April 2013.
[2] Adediran Y.A, Reyaz T A “Comparative analysis of modulation Techniques in mobile communication systems”,
proceedings of conference on GSM in Nigeria, 2003.
[3] Wireless Link Budget Analysis whitepaper, Tranzeo wireless technologies inc.
[4] B. Razavi, RF Microelectronics, Prentice-Hall, 1998.
[5] Fundamentals of RF and Microwave Noise Figure Measurements Agilent Technologies.
[6] Aggarwal .V “Analog Circuit Optimization using evolutionary algorithms and convex optimization”, MSc Thesis:
Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology. May 2007.
[7] Vancorenland. P, De Ranter. C., Steyaert. M., Gielen .G. “Optimal RF design using smart evolutionary algorithms”.
sigda.org/Archives/.../Dac/Dac2000/papers/2000/dac00/pdffiles/01_2.pdf source date:2nd October 2008.
[8] Barros M, Guilherme.J, Horta. N.”Optimization and synthesis of analog circuit and system using evolutionary
algorithm techniques”, Instituto Politecnico de lomar. Portugal 2007.
[9] Qi Z, Ziegler .M, Kosonocky .S.V, Rabaey.J.M, Mircea R.S “Multi dimensional circuit and micro-architecture level
optimization”, Proceedings of the 8th International symposium on quality electronic design (ISQED„07).IEEE
computer society. 2007.