2019 8th MEDITERRANEAN CONFERENCE ON EMBEDDED COMPUTING (MECO), 10-14 JUNE 2019, BUDVA, MONTENEGRO

Simulation of System with Code Channel Division in

the LabVIEW

I.S. Trubin
Vyatka State University, VyatSU
Kirov, Russia
[email protected]

Abstract— Model of digital communication system with code (CDMA) in the LabVIEW environment compatible with the NI
channel division in LabVIEW is developed. During the USRP SDR system [4] and test the performance of the model
development of the model, the focus was on performance and when receiving a real signal. In Fig. 1 shows the block diagram
compatibility with the SDR (Soft Defined Radio) NI USRP-2920. of the transmitter, and Fig. 2, respectively, of the receiver.
The network level of the system is replaced by a source of
random bits that form an analogue of speech frames. Using II. TRANSMITTER MODEL
modeling in LabVIEW, the dependence of the probability of the
appearance of distorted information frames at the output of the
Let (l + 1)-th network layer frame in the i-th traffic channel
Viterbi decoder – FER (Frame Error Ratio) on the signal-to- is given by a polynomial
noise ratio (SNR) at the decoder input was obtained. 171(79)
(i )
ul +1 ( D ) = ∑ u (ji+)172(80)l D j , (1)
Keywords-CDMA, simulation, channel and physical levels, j =0
SDR, LabVIEW
where D - delay by 1 step.
I. INTRODUCTION A continuous stream of bits in the i-th channel will be set
Information transfer technology with code division of by a polynomial
channels – CDMA (Code Division Multiple Access) has been
∞ ( i ) 12 8 (i )
actively developed and used since the 50s of the twentieth u ( ) ( D ) ∑ (ul D ( ) + rl ( D)) D
=
i 8+192( 96 )l
. (2)
century, in military and civil communication systems [1]. The l =0
broad development of communication systems with code
division of channels a long time was constrained by complexity (i )
The remainder rl ( D) is given by the expression
of program and hardware realization of necessary mathematical
algorithms of processing of signals. The advent of SDR (i )
technology has significantly simplified the implementation of rl ( D ) = ul(i ) D12(8)    
mod g12(8) ( D ) , (3)
new digital signal processing technologies. SDR allowed to
create various radio-technical systems with the help of where g12(8) ( D ) - convolutional code generator.
software-configurable transceivers. The developer has been Formula (2) describes the process of adding a cyclic prefix
able to implement complex algorithms of digital signal to a network layer frame in order to control the decoding
processing in software, without thinking about their circuit accuracy at the receiving side. This operation is performed by
implementation [2]. the CRC-12 unit (Fig. 1)
Specialized software products such as LabVIEW, A convolutional encoder forms a bitstream defined by a
MATLAB (Simulink), GNU Radio and others are used to polynomial
reduce the complexity of developing complex digital signal
processing systems in SDR [3]. However, even in commercial
( ) ( )
∞ 2 ∞
j −1 ( i )
systems, individual units that model certain digital signal v ( D)
= =
∑ v '(i−)1 2
∑ ∑gj D D u D2 . (4)
j= 0 f ( j) j= 1l= 0
processing operations do not always satisfy developers in terms
of speed and accuracy of the implementation of individual The right side of (4) is performed by the Convolution
functional operations of modern communication standards. encoder block (Fig. 1). In (4) f (k ) sets the permutation of the
Accordingly, the creation of receiver and transmitter models block interleaver
for a system that implements CDMA compatible with SDR is
an important task. =f (k ) ( k   24
mod ) *16 + k    24
div . (5)
The purpose of this work was to develop a model of a
.
digital communication system with code channel division

978-1-7281-1740-9/19/$31.00 ©2019 IEEE

 2019 8th MEDITERRANEAN CONFERENCE ON EMBEDDED COMPUTING (MECO), 10-14 JUNE 2019, BUDVA, MONTENEGRO

Encoding
pi Walsh-Hadamard Sequential bit
Upsamp I
●
●
● transformation ● reading ling
FIR
● ● 0 63
● ● ●
CRC-12 Conv. Repeat Interlea 0
I ch encoder encoder symbol ver I-pilot sequence
● ● ●
Layer 3 ● ● ● DAC
LFSRs_IQ
● ●●●●● ●●●●●●● ● ●
● ● ●
● ● ● 383
● ● ● Q-pilot sequence
● CRC-12 Conv. Repeat Interlea ● ●
Upsamp Q
63 ch encoder encoder symbol ver FIR
ling

Figure 1. Block diagram of the transmitter.

Figure 2. Block diagram of the receiver.

The process of formation of the complex envelope at the 

transmitter output reflects the formula 

∞ ∞  
=s (t ) ∫  ∑ ai I i   2
mod 16 δ ( τ − iTb )  h ( t − τ ) d τ , (6) 
4α
−∞  i = 0   1− α + , α = 0, 22
 π
where I i - element of scrambling pseudo-random sequence.  (7)
 α  2   π   2   π  T
h (t ) =  1 + π  sin  4α  + 1 − π  cos  4α   , t = ± c
Next, the signal is scrambled (multipliers and LFSRs block,  2         4α
Fig. 1. The impulse response of the shaping filter is given by   t t  t  
(7) (signal like root raised cosine).  sin  π (1 − α ) + α cos  π (1 + α )  
  Tc Tc  Tc 
 .
 2
 t   4αt  
 π 1−  
 Tc   Tc  
 
 2019 8th MEDITERRANEAN CONFERENCE ON EMBEDDED COMPUTING (MECO), 10-14 JUNE 2019, BUDVA, MONTENEGRO

In Fig. 5 shows the dependence of the frequency on the

appearance of distorted information frames FER (Frame Error
Ratio) on the output of the Viterbi decoder from the
signal/noise ratio (SNR) at the decoder input received by the
simulation in LabVIEW.
As can be seen from Fig. 5, the fraction of distorted frames
rapidly decreases from an unacceptable level (~ 0.1) to a
negligible ((~ 0.001) in a small interval of 2-3.9 dB. In the
simulation, maximum noise immunity was obtained for a
convolutional code with a code limit length K c = 8 and a code
Figure 3. The signal power spectrum at the transmitter output (LabVIEW).
1
rate Rc = .
Because LabVIEW operates, only discrete values [2] 2
further assume that all these formulas are implemented in B. WCDMA receiver model
discrete time.
For implementation WCDMA standard receiver in
A pulse of the form (7) determines a band-pass signal LabVIEW, let us represent the work of the receiver as some
without intersymbol interference with fast attenuation finite state machine
1
asymptotic proportional to 2 . The signal power spectrum at only only !test _ work ()
t def df → start → work → def df .
the transmitter output is shown in Fig. 3.
Discrete convolution (6) is performed using the
Upsampling and FIR blocks (Fig. 1).
When performing discrete convolution some samples will
be obviously zero, this allows reducing the number of
operations performed. The described implementation of the
transmitter is capable of operating in real time.
III. RECEIVER MODEL
A. RAKE receiver model
Let us set the impulse response of the communication
channel by the expression
L
hCH ( t )= ∑ αl e jϕl δ ( t − τl ) . (8)
l =1
Figure 4. The receiver elements in LabVIEW.
The formula (8) takes into account the presence of reflected
signals on the receiving side. We have implemented a model of
Rake-receiver (Fig. 2 a), which consists of synchronous
detector envelope, "comb" correlators and combinatorial logic.
At the output, we get a "soft" solution
To increase the speed when sampling the envelope at the
output of the USRP receiver, two samples are also used per bit
interval Tb , which leads to a blurring of the orthogonality
Tb
effect. The maximum temporal uncertainty can be .
4
The implementation of the beam search unit (Fig. 2c) does
not take into account the possibility of signals from several
base stations, and all detected rays are fed to a Rake receiver,
although some rays must be discarded due to their low energy.
Figure 5. Dependence of the frequency on the appearance of distorted
The implementation of the interleaver in LabVIEW is information frames FER (Frame Error Ratio) on the output of the Viterbi
shown in Fig. 4. The interleaver and FER block are provided decoder from the signal/noise ratio (SNR) at the decoder input.
the achievement and control of the state of personnel
synchronism.
 2019 8th MEDITERRANEAN CONFERENCE ON EMBEDDED COMPUTING (MECO), 10-14 JUNE 2019, BUDVA, MONTENEGRO

Figure 6. The interaction of software models and USRP.

In an initial condition of def df the rough estimate of of appearance of distorted frames due to clock sync uncertainty
frequency shift of ∆f of a signal is given (with a step of 10 Hz), and low sampling rate.
the scrambling code is defined. The condition of start is
initialized by the resempler who is able work decoding of a V. CONCLUSION
signal is carried out. At detection of failure of synchronism In this work, the models of the transmitter and receiver of
transition to def df happens the test_work() function. CDMA realized. The models implement all the basic
Next, based on the blocks that implement the basic signal operations performed on the signal, in accordance with the
processing operations, we have developed a system-level functional model of data transmission with code division of the
model of the standard WCDMA receiver in LabVIEW. channel.
When developing individual modules of models, the main
IV. RESULTS attention was paid to the speed and compatibility with the SDR
For experimental verification of the presented models, we platform NI USRP. The performance check of the developed
used two USRP-2920. The interaction of software models and models was carried out using NI USRP-2920.
USRP is shown in Fig.6. Significant number arrays of
constants: permuting the code; reference signals of the The network level of the system is replaced by a source of
random bits that form an analogue of speech frames. To control
correlators; the coefficients of forming filters Nyquist, formed
with the auxiliary virtual instruments from the library of the the receiver model in LabVIEW, we developed a virtual
LabVIEW Modulation toolkit are used in the developed instrument with a software interface.
software models. Using modeling in LabVIEW, the dependence of the
The network level of the system was replaced by a source probability of the appearance of distorted information frames at
of random bits that form an analogue of speech frames the output of the Viterbi FER (Frame Error Ratio) decoder on
(“Source of information frames” in Fig. 6). the signal-to-noise ratio (SNR) at the decoder input was
obtained. The WCDMA receiver model allows you to select
The simulation showed that when USRP sampling rate CPICH and CCPCH channels in a real signal received from a
drops to 850000 samples per second, it is possible to enter real- satellite.
time mode. The processing speed is approximately 17 frames
per second.
REFERENCES
At a higher sampling rate, a USRP buffer (or RAM)
[1] H. Harada and R. Prasad, Simulation and Software Radio for Mobile
overflow occurs and an overflow error occurs.
Communications, Artech House: 2002.
We produced several experiments at different distances [2] N..Kehtarnavaz and N. Kim, Digital signal processing system-level
between the receiver and transmitter. The maximum distance design using LabVIEW,Dodeka XXI vek,2007 (in Russian).
corresponds to 50 meters. A model in LabVIEW using the [3] L.C. Clark, Digital Signal Processing: and Digital Communication,
simplest model of a multipath communication channel showed McGraw-Hill Education; 1 edition, 2005.
better results than an experiment with USRP in the frequency [4] USRP-2920 Specifications - National Instruments [Electronic resource],
URL: http://www.ni.com/pdf/manuals/375839c.pdf (04.04.2019).