Modeling and Refining Heterogeneous Systems With System CAMS

  • Slides: 22
Download presentation
Modeling and Refining Heterogeneous Systems With System. C-AMS: Application to WSN M. Vasilevski F.

Modeling and Refining Heterogeneous Systems With System. C-AMS: Application to WSN M. Vasilevski F. Pecheux, N. Beilleau, H. Aboushady K. Einwich* Laboratory LIP 6 University Pierre and Marie Curie, Paris 6, France *Fraunhofer IIS/EAS, Dresden, Germany M. Vasilevski March University 2008 Paris 6 Fraunhofer IIS/EAS

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description c. SDF Multi-rates 3. RF and AMS Modeling a. AMS Models b. RF Models 4. Wireless Sensor Network Node 5. Conclusion M. Vasilevski University Paris 6 Fraunhofer IIS/EAS

1. Issues : Mixed Systems Design Matlab System. C Matlab Verilog-A VHDL-AMS Verilog VHDL

1. Issues : Mixed Systems Design Matlab System. C Matlab Verilog-A VHDL-AMS Verilog VHDL Verilog-A VHDL-AMS Spice A/D Converter M. Vasilevski Spice-RF Microcontroller University Paris 6 Fraunhofer IIS/EAS RF Transceiver 3

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description c. SDF Multi-rates 3. RF and AMS Modeling a. AMS Models b. RF Models 4. Wireless Sensor Network Node 5. Conclusion M. Vasilevski University Paris 6 Fraunhofer IIS/EAS

2. a Models of Computation System. C-AMS Synchronous Data Flow SDF Modeling Formalism System.

2. a Models of Computation System. C-AMS Synchronous Data Flow SDF Modeling Formalism System. C Linear Network LN Modeling Formalism LN Solver Models of computation : Other Modeling Formalism Other Solver DE, Mo. Cs (CP, FSM, etc…) Synchronisation Layer • Conservative Linear network • Synchronous Data Flow System. C Simulation Kernel M. Vasilevski University Paris 6 Fraunhofer IIS/EAS 5

2. b SDF Behavioral Description SCA_SDF_MODULE(B) B Input Behaviour Output SCA_SDF_IN<double> SCA_SDF_OUT<double> void sig_proc(

2. b SDF Behavioral Description SCA_SDF_MODULE(B) B Input Behaviour Output SCA_SDF_IN<double> SCA_SDF_OUT<double> void sig_proc( ) A M. Vasilevski C University Paris 6 Fraunhofer IIS/EAS 6

2. c SDF Multi-Rates Cluster Simulation sample time Tin A 1 2 1 Tout

2. c SDF Multi-Rates Cluster Simulation sample time Tin A 1 2 1 Tout B 16 k. Hz 8 Hz 3 2 48 k. Hz C 1 24 k. Hz Simulation rates M. Vasilevski University Paris 6 Fraunhofer IIS/EAS 7

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description c. SDF Multi-rates 3. RF and AMS Modeling a. AMS Models b. RF Models 4. Wireless Sensor Network Node 5. Conclusion M. Vasilevski University Paris 6 Fraunhofer IIS/EAS

3. a AMS models : Integrator In/Out ports Other Attributes Initialisation method Signal processing

3. a AMS models : Integrator In/Out ports Other Attributes Initialisation method Signal processing method M. Vasilevski SCA_SDF_MODULE (integrator) { sca_sdf_in < double >in; sca_sdf_out < double >out; double f; sca_vector < double >NUM, DEN, S; sca_ltf_nd ltf 1; void set_coeffs(double A){ DEN (0) = 0. 0; DEN (1) = 1. 0; NUM (0) = A; } void sig_proc(){ out. write( ltf 1(NUM, DEN, S, in. read())); } SCA_CTOR (integrator) {}}; University Paris 6 Fraunhofer IIS/EAS 9

3. a AMS models : Decimator SCA_SDF_MODULE (decimator) { sca_sdf_in < double >in; sca_sdf_out

3. a AMS models : Decimator SCA_SDF_MODULE (decimator) { sca_sdf_in < double >in; sca_sdf_out < double >out; double old_input; 2 M. Vasilevski void init(){ in. set_rate(2); out. set_rate(1); old_input=0; } 2 void sig_proc(){2 double input=in. read(0)/2; out. write(old_input+input); old_input=input; } SCA_CTOR (decimator){} }; University Paris 6 Fraunhofer IIS/EAS 10

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description c. SDF Multi-rates 3. RF and AMS Modeling a. AMS Models b. RF Models 4. Wireless Sensor Network Node 5. Conclusion M. Vasilevski University Paris 6 Fraunhofer IIS/EAS

3. b RF models Power gain IIP 3 NF Rin Rout Na input a

3. b RF models Power gain IIP 3 NF Rin Rout Na input a 1 x+a 3 x³ Rin Rout output a 1 = f(Power gain, Rout) a 3 = f(a 1, IIP 3) Na = f(NF) M. Vasilevski University Paris 6 Fraunhofer IIS/EAS 12

3. b RF models : IIP 3 and Noise Figure Test FFT BW =

3. b RF models : IIP 3 and Noise Figure Test FFT BW = 120 k. Hz Power Gain = 10 d. B Input amplitude = -16. 02 d. Bm IIP 3 = 10 d. Bm NF = 30 d. B M. Vasilevski University Paris 6 Fraunhofer IIS/EAS 13

3. b RF models : Baseband Equivalent X(t) = DC + I 1 cos(wt)

3. b RF models : Baseband Equivalent X(t) = DC + I 1 cos(wt) + I 2 cos(2 wt) + I 3 cos(3 wt) + Q 1 cos(wt) + Q 2 cos(2 wt) + Q 2 cos(3 wt) DC I 1 0 Q 1 M. Vasilevski w Q 2 I 3 x. BB(t) = 2 w 3 w Q 3 University Paris 6 Fraunhofer IIS/EAS 14

3. b RF models : Baseband Equivalent Implementation class BB{ double DC, I 1,

3. b RF models : Baseband Equivalent Implementation class BB{ double DC, I 1, I 2, I 3, Q 1, Q 2, Q 3; . . . BB operator+(BB x)const{ SCA_SDF_MODULE (adder) { BB z(this->DC+x. DC, sca_sdf_in < BB double >in. I; this->I 1+x. I 1, sca_sdf_in < BB double >in. Q; this->I 2+x. I 2, sca_sdf_out < BB double >out; this->I 3+x. I 3, . . . this->Q 1+x. Q 1, void sig_proc () { this->Q 2+x. Q 2, out. write (in. I. read()+ this->Q 3+x. Q 3); in. Q. read()); return z; }. . . University Paris 6 15 }; M. Vasilevski Fraunhofer IIS/EAS

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description

1. Issues 2. System. C-AMS Language a. Models of Computation b. SDF Behavioral Description c. SDF Multi-rates 3. RF and AMS Modeling a. AMS Models b. RF Models 4. Wireless Sensor Network Node 5. Conclusion M. Vasilevski University Paris 6 Fraunhofer IIS/EAS

4. Wireless Sensor Network Node • Wireless sensor network for environmental and physical monitoring:

4. Wireless Sensor Network Node • Wireless sensor network for environmental and physical monitoring: o Temperature, vibration, pressure, motion, polluants M. Vasilevski University Paris 6 Fraunhofer IIS/EAS 17

4. Wireless Sensor Network Node System. C-AMS System. C A/D Converter ATMEGA 128 8

4. Wireless Sensor Network Node System. C-AMS System. C A/D Converter ATMEGA 128 8 bits Microcontroller modulator 2 nd order Application OSR=64 Binary File 10 bits decimator RZ feedback 8. 53 MHz M. Vasilevski 2. 4 MHz University Paris 6 Fraunhofer IIS/EAS RF Transceiver QPSK fc=2. 4 GHz 2. 4 GHz 18

4. Wireless Sensor Network Node RF : QPSK 2. 4 GHz filter encoder mux

4. Wireless Sensor Network Node RF : QPSK 2. 4 GHz filter encoder mux LNA demux filter ADC : + - + decimator DAC M. Vasilevski University Paris 6 Fraunhofer IIS/EAS 19

4. Wireless Sensor Network Node : Results Noisy channel DC offset RF Simulation (2.

4. Wireless Sensor Network Node : Results Noisy channel DC offset RF Simulation (2. 4 GHz) SC-AMS classical simulation SC-AMS BB eq. RF simulation 1000 bits 63. 0 s transmission 0. 036 s DC offset 19. 9 s 0. 018 s Frequency offset 24. 9 s 0. 022 s Phase mismatch 44. 4 s 0. 031 s M. Vasilevski Frequency offset University Paris 6 Fraunhofer IIS/EAS Phase mismatch 20

4. Wireless Sensor Network Node : Results Settings Simulation ADC alone Matlab System. C-AMS

4. Wireless Sensor Network Node : Results Settings Simulation ADC alone Matlab System. C-AMS OSR=64 16*1024 pts 1. 6 s 10 bits 8. 53 MHz 0. 9 s RF alone 2. 4 GHz 10 e 3 bits 150. 7 s classic BB 10 e 7 pts RF 63. 0 s 0. 036 s 2 -nodes Same 10 e 3 bits 181. 7 s transmission settings M. Vasilevski University Paris 6 Fraunhofer IIS/EAS 21

Conclusion Advantages to use System. C-AMS: • Digital and Analog-Mixed Signal systems simulation §

Conclusion Advantages to use System. C-AMS: • Digital and Analog-Mixed Signal systems simulation § Interface with System. C • Simulations very fast § C++ based • Polymorphism § Easy to refine components with C++ inheritance ability § Generic declaration of components with C++ templates • Easy software programmer contribution § Example of a free FFT library used for IIP 3 test. M. Vasilevski University Paris 6 Fraunhofer IIS/EAS 22

Modeling Heterogeneous Systems Using System CAMS A CaseModeling Heterogeneous Systems Using System CAMS A Case
Cams Cams are used to convert rotary motionCams Cams are used to convert rotary motion
Lecture 20 CAMs ROMs PLAs 20 CAMs ROMsLecture 20 CAMs ROMs PLAs 20 CAMs ROMs
Dynamic Mechanisms CAMS Cams A cam is aDynamic Mechanisms CAMS Cams A cam is a
Heterogeneous Homogeneous Solution Mixtures Solutions Heterogeneous colloid HeterogeneousHeterogeneous Homogeneous Solution Mixtures Solutions Heterogeneous colloid Heterogeneous
The Refining Industry in Iraq Present Installed RefiningThe Refining Industry in Iraq Present Installed Refining
Projecting Financial Statements REFINING OUR ASSUMPTIONS Refining ModelProjecting Financial Statements REFINING OUR ASSUMPTIONS Refining Model
REFINING FUNDAMENTALS Presentation Outline Crude Oil Characteristics RefiningREFINING FUNDAMENTALS Presentation Outline Crude Oil Characteristics Refining
Petroleum Refining Process Overview of petroleum refining processPetroleum Refining Process Overview of petroleum refining process
PS Heterogeneous Systems Handson Acceleration on Heterogeneous ComputingPS Heterogeneous Systems Handson Acceleration on Heterogeneous Computing
Lesson 17 Heterogeneous and cloud processes Heterogeneous andLesson 17 Heterogeneous and cloud processes Heterogeneous and
Catalysis and Heterogeneous Catalysis Catalysis and Heterogeneous CatalysisCatalysis and Heterogeneous Catalysis Catalysis and Heterogeneous Catalysis
Homogeneous and heterogeneous catalysis Heterogeneous catalysis This projectHomogeneous and heterogeneous catalysis Heterogeneous catalysis This project
12 7 NOTES Heterogeneous Mixtures IV Heterogeneous Mixtures12 7 NOTES Heterogeneous Mixtures IV Heterogeneous Mixtures
Heterogeneous datacenter interop solutions Reducing Complexity in HeterogeneousHeterogeneous datacenter interop solutions Reducing Complexity in Heterogeneous
Managing Heterogeneous Audience Heterogeneous Audience It is definesManaging Heterogeneous Audience Heterogeneous Audience It is defines
Heterogeneous Processing KYLE ADAMSKI Overview What is heterogeneousHeterogeneous Processing KYLE ADAMSKI Overview What is heterogeneous
Heterogeneous catalysis Basic principles of heterogeneous catalysis ThisHeterogeneous catalysis Basic principles of heterogeneous catalysis This
Outline Introduction to Heterogeneous Networks Het Net HeterogeneousOutline Introduction to Heterogeneous Networks Het Net Heterogeneous
Outline Introduction to Heterogeneous Networks Het Net HeterogeneousOutline Introduction to Heterogeneous Networks Het Net Heterogeneous
METO 621 CHEM Lesson 3 Heterogeneous Chemistry HeterogeneousMETO 621 CHEM Lesson 3 Heterogeneous Chemistry Heterogeneous