modeling wireless sensor network nodes using systemc-ams

Modeling Wireless Sensor Network Nodes Using SystemC-AMS

Modeling Wireless Sensor Network Nodes

Using SystemC-AMS

Michel Vasilevski, Hassan Aboushady,Francois Pecheux, Laurent de [email protected]

Laboratory LIP6

4, place Jussieu

75252 PARIS Cedex 05

December 2007


Introduction

SystemC-AMS introduction

Synchronous DataFlowConservative model

Wireless sensor network node : Architecture and implementation

Analog-to-Digital ConverterArchitecture

Implementation

MicrocontrollerQPSK RF Transceiver

Transmitter architecture

Receiver architecture

RF Validation

Conclusion


Introduction

Introduction

Wireless sensor networkfor environmental andphysical monitoring:

Temperature, vibration, pressure,motion, poluants

Analog-digital mixed

systems description

needed




LNModelingFormalism

LNSolver

OtherModelingFormalism

OtherSolver

ModelingSDF

Formalism

DE MoCs

(CP, FSM,etc...)

Synchronisation Layer

SystemC Simulation Kernel

Models of computation(MoC)

Model : SynchronousData Flow (SDF)

Model : Conservativelinear network



Synchronous DataFlow

Synchronous DataFlow : multi-rate synchronisation

port sample time = 1/16e3

= 62.5µs

2 1 3 2B

out sample freq

out sample rate= in sample freq

in sample rate

1 1

8 kHz 16 kHz 24 kHz48 kHz

port rates

A C



Synchronous DataFlow

Synchronous DataFlow : language elements

SCA_SDF_MODULE (interpolator)

sca_sdf_in <double> in; // input port

sca_sdf_out <double> out; // output port

...

SCA_CTOR(interpolator)...

void init()

in.set_T(62.5,SC_US); // simulation sample time setting

// freq = 16 kHz

in.set_rate(1); // 1 sample consumed

out.set_rate(3); // 3 samples produced

...

void sig_proc()

double input = in.read();

for (int i=0;i<3;i++) out.write(input,i);

;



Conservative model

Conservative model

sca r

sca c

sca l

sca vconst / sca vsin

sca iconst / sca isin



Conservative model

Conservative model : language elements

1 SC_MODULE (wave )

2

3 sca_elec_port w1;

4 sca_elec_ref gnd;

5

6 sca_isin *i_sin_1;

7 sca_r *i_r_1 ;

8

9 void init (double a, double f)

10 i_r_1 ->value = a*1000;

11 i_sin_1 ->freq = f;

12

13 SC_CTOR (wave )

14 i_sin_1=new sca_isin (" i_sin_1 ");

15 i_sin_1 ->p(w1);

16 i_sin_1 ->n(gnd);

17 i_sin_1 ->ampl =0.001;

18

19 i_r_1 =new sca_r ("r1");

20 i_r_1 ->p(w1);

21 i_r_1 ->n(gnd );

22

23 ;

R = xkΩ

u(t) =

i(t) =sin(2πft)mA

xsin(2πft)V



Wireless sensor network node : Architecture and

implementation

SystemC

SystemC-AMSSDF modelConservative model

SystemC-AMS

2.4 GHz2.4 MHz8.53 MHz — 133 kHz3.65 kHz

Binary FileDecimator

modulatorΣ∆ RF Transceiver

8 bits microcontrollerATMEGA128

Application

A/D Conversion

Sensor



Analog-to-Digital Converter


1 bit

int1 int2

1a

A A

a2

Ts

X(s) Y(z)

DAC

U(z)

Ts

H (s)

T

x[n] y[n]2 2 2(1+ z−1)k (1+ z−1)k (1+ z−1)k

Modulator Σ∆:

resolution 10bits

2nd order

OSR=64

Return-to-Zerofeedback

Decimator:

FIR2 filter

3rd order(k=3)




Implementation : SDF model of the Σ∆ integrator

1 SCA_SDF_MODULE (integrator_sd )

2

3 sca_sdf_in < double >in;

4 sca_sdf_out < double >out;

5

6 double fs, ai;

7 sca_vector < double >NUM , DEN , S;

8 sca_ltf_nd ltf1 ;

9

10 void init (double a, double f)

11 DEN (0) = 0.0;

12 DEN (1) = 1.0;

13 NUM (0) = 1.0/3.0;

14 ai=a;

15 fs=f;

16

17 void sig_proc ()

18 out.write (ltf1 (NUM ,

19 DEN ,

20 S,

21 fs * ai * in.read ()

22 )

23 );

24

25 SCA_CTOR (integrator_sd )

26 ;




Validation : Σ∆ simulation

SNR variation repecting to input amplitude

Power spectrum of output signals

−60 −50 −40 −30 −20 −10 010

20

30

40

50

60

70

Amplitude (dB)

SN

R (d

B)

RZ : matlabRZ : SystemC−AMS

SN

R(d

B)

Matlab

SystemC-AMS

Input amplitude (dB)

100

101

102

103

104

−140

−120

−100

−80

−60

−40

−20

0

100

101

102

103

104

−150

−100

−50

0

Normalized Frequency, f/(fs/64)

Po

we

r S

pe

ctra

l De

nsi

ty (

dB

)

MatlabSystemC−AMS simulation

MatlabSystemC−AMS simulation

Modulator output

Decimator output



Microcontroller

Microcontroller

ATMEGA128 2.4 MHz

CABA : Cycle Accurate, Bit Accurate

Binary FileBinary File

ATMEGA128

Sensor

RF Transceiver

Sensor

ApplicationApplication

ATMEGA1288 bits microcontroller RF Transceiver

A/D Conversion A/D Conversion

8 bits microcontroller



QPSK RF Transceiver

QPSK RF Transmitter

QPSK : quadrature phase shift keying

0 100 200 300 400 500 600−1

−0.8

−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

1

0 100 200 300 400 500 600−1

−0.8

−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

1

0 100 200 300 400 500 600−1.5

−1

−0.5

0

0.5

1

1.5

−√

Eb

+√

Eb

Demultiplexer

φ2(t) =√

2T

sin(2πfct)

φ1(t) =√

2T

cos(2πfct)

Binary

Data

−√

Eb,√

Eb,√

Eb...Encoder

011011...

antenna

sQ

sI

0

1

−√

Eb

+√

Eb +√

Eb

−√

Eb



QPSK RF Transceiver

QPSK RF Receiver

0

1

0 100 200 300 400 500 600−1.5

−1

−0.5

0

0.5

1

1.5

0 100 200 300 400 500 600−1.5

−1

−0.5

0

0.5

1

1.5

0 100 200 300 400 500 600−1

−0.8

−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

1

0 100 200 300 400 500 600−1.5

−1

−0.5

0

0.5

1

1.5

0 100 200 300 400 500 600−1

−0.8

−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

1

1

0

0

1

φ1(t) =√

2T

cos(2πfct)

φ2(t) =√

2T

sin(2πfct)

BinaryData

011011 . . .

DecisionDevice

Multiplexer

1T

1T

DecisionDevice

threshold = 0

threshold = 0

antenna Filter

Filter



QPSK RF Transceiver

QPSK RF Receiver

0

1

0 100 200 300 400 500 600−1.5

−1

−0.5

0

0.5

1

1.5

0 100 200 300 400 500 600−1.5

−1

−0.5

0

0.5

1

1.5

0 100 200 300 400 500 600−1

−0.8

−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

1

0 100 200 300 400 500 600−1.5

−1

−0.5

0

0.5

1

1.5

0 100 200 300 400 500 600−1

−0.8

−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

1

0

1

1

−1.5 −1 −0.5 0 0.5 1 1.5−1.5

−1

−0.5

0

0.5

1

1.5Constellation de la transmission idéale

φ1(t) =√

2T

cos(2πfct)

φ2(t) =√

2T

sin(2πfct)

1T

threshold = 0

antenna

1T

DeviceDecision

Filter

Filter

Constellation



QPSK RF Transceiver

RF Validation : Simulation including effects of

defaults

sin(2π(fc +∆f )t)

cos(2π(fc +∆f )t)

sin(2πfct −Φ/2)

cos(2πfct +Φ/2)

Offset

Offsetsin(2πfct)

cos(2πfct)

sin(2πfct)

cos(2πfct)

AWGN

DC Offset

Phase mismatchFrequency offset

Noisy channel



QPSK RF Transceiver

RF Validation : BER variation respecting to SNR

0 1 2 3 4 5 6 7

10−3

10−2

10−1

SNR (dB)

BE

R

TheorySystemC−AMSTheory

SystemC-AMS

SNR (dB)

BER

BER =nb bits incorrect

nb bits sent



QPSK RF Transceiver

Simulation results

Settings Simulation Matlab SystemC-AMS

OSR=64 1 msADC alone 8 bits output 16*1024 pts 1.605s 0.934s

416.67 µs

RF alone 2.4 GHz 103 bits 2m30.746s 54.360s107 pts RF

2-mote SameWSN settings 416.67 µs – 3m1.654s


Conclusion

Conclusion

SystemC

SystemC-AMSSDF modelConservative model

SystemC-AMS

2.4 GHz2.4 MHz8.53 MHz — 133 kHz3.65 kHz

Binary FileDecimator

modulatorΣ∆ RF Transceiver

8 bits microcontrollerATMEGA128

Application

A/D Conversion

Sensor

Ongoing research :

Refinements : IIP3, CP1, Noise Figure, I/O Impedances

Baseband equivalent modeling

modeling wireless sensor network nodes using systemc-ams

Documents