In the course of the ANDRES project, a Library of configurable SystemC AMS modules was developed at the Vienna University of Technology, and is . The implemented building blocks focus on the field of communication and radio frequency systems, in particular on signal sources, modulation/demodulation blocks, filters, measurement and observation parts. They can be subdivided into three categories of blocks:

• Signal sources
• Basic blocks for signal processing
• Signal analysis units

The library also contains some example application. One example, an OFDM transceiver, is described here.

Download the SystemC AMS Building Block Library here!

Examples (Full documentation available with package)

Gaussian distributed random Numbers 

Functional description
This module generates Gaussian distributed random numbers. The parameters “_mean” and “_variance” are used to set the mean and variance of the random numbers. The data rate of the output port can be increased with the parameter “_rate”.

Module definition:    noise_gauss(sc_module_name nm, double _variance, double _mean, int _rate);
Module Interface:     sca_tdf::sca_out<double> out;

Parameters:

Low-noise amplifier (LNA)

Functional description
This module amplifies an input signal with a certain gain. User can set the gain with the parameter “_gain”. The non ideal effect “intercept modulation” can also be modelled with this module if necessary. With the parameter “_ideal” the user can switch between an ideal and a non-ideal LNA module. The IP3 point can be set with the parameter “_ip3”.

Module definition:    lna(sc_module_name n, double _gain, double _ip3, bool _ideal);
Module interfaces:    sca_tdf::sca_in<double> in;   
                                   sca_tdf::sca_out<double> out; 

Parameters:

OFDM modulator (OFDM transmitter)

Functional description
This module modulates a bit stream to an OFDM carrier signal, which is the sum of a number of orthogonal sub-carriers, with baseband data on each sub-carrier being independently modulated using quadrature amplitude modulation (QAM). The bit width of symbol can be set with the template parameter N.

Module definition:    template <int N>
ofdm_se(sc_module_name n,int qam_p_num,double mixer_fc,double bit_f,int dout_rate,double _amp)
Module interfaces:    sca_tdf::sca_in<bool> in;   
                                   sca_tdf::sca_out<double> out; 
Parameters:

int sc_main(int argc, char* argv[])
{
  sc_set_time_resolution(1, SC_PS);	// set the time resolution

  sca_tdf::sca_signal<bool> sig_stimuli;    // declare SystemC AMS signals
  sca_tdf::sca_signal<double> sig_out;
  sca_tdf::sca_signal<double> noise_out;
  ...

  rand_bool i_stimuli("stimuli",16);   // instantiate test generator
  ofdm_se<8> i_tran("transmitter",constl_dim,
                    freq_carrier,freq_bit,data_rate,ampl_se); // instantiate transmitter
  air i_air("air",attent,"gauss_white",n_va,n_mean,data_rate);
						// instantiate channel
  ofdm_re<8> i_receiver("receiver",constl_dim,freq_carrier,freq_bit,data_rate, 
1ampl_re);			// instantiate receiver
  drain drn("drn");                    // instantiate drain. This is only a 
// module used to consume tokens as we       
// can not let a SystemC AMS scheduling 
// loop open.

  i_stimuli.out.set_timestep(1/freq_bit,SC_SEC); //Time step has to be set to one of 
      the ports in the system

  i_stimuli.out(sig_stimuli);          // connect different modules
  i_tran.in(sig_stimuli);
  i_tran.out(sig_out);  
  ...

  // Tracing:
  sca_util::sca_trace_file wave = sca_util::sca_create_vcd_trace_file ("wave"); 
  sca_util::sca_trace (wave, sig_stimuli ,"stimuli");
  ...

  sc_start(0.05,SC_MS);                // Simulation time: 0.05ms  
  return 0;
}

The transmitter takes a serial stream of binary digits. By inverse multiplexing, these are first demultiplexed into N parallel streams, and each one mapped to a (possibly complex) symbol stream using QAM modulation.

An inverse FFT is computed on each set of symbols, giving a set of complex time-domain samples. These samples are then quadrature-mixed to passband in the standard way.

Block diagram of OFDM transmitter

The receiver picks up the signal from antenna, which is then quadrature-mixed down to baseband using cosine and sine waves at the carrier frequency. This also creates signals centered on 2fc, so low-pass filters are used to reject these. The baseband signals are then sampled and a forward FFT is used to convert back to the frequency domain. This returns N parallel streams, each of which is converted to a binary stream using an appropriate symbol detector. These streams are then re-combined into a serial stream, which is an estimate of the original binary stream at the transmitter.

Block diagram of OFDM Receiver

The next code Section shows how to use building block modules to build the OFDM transmitter module:

template <int N>
SC_MODULE(ofdm_se){
public:
 //Ports:
 sca_tdf::sca_in<bool> in;            // declare port
 sca_tdf::sca_out<double> out;        
 //signals
 sca_tdf::sca_signal<bool> sig_pa[N];  	// declare signals to 
 sca_tdf::sca_signal<double> sig_real[N];  // connect sub modules
 ...
private:
  //module instantiation
  s2p<bool,N>* s2p_sub;
  qam_map* qam_mapper_sub[N];  
  fft_ifft<N>* ifft_sub;
  p2s<double,N>* p2s_r_sub;
  p2s<double,N>* p2s_i_sub;
  q_mixer_tr* q_mixer_tr_sub; 
 public:
  //Constructor
  ofdm_se(sc_module_name n,int qam_p_num,double mixer_fc,double bit_f,int
   dout_rate,double _amp=1)
{  
   int mixer_rate=(int)floor(dout_rate*log2(qam_p_num)*mixer_fc/bit_f);
// instantiating modules and connecting them using signals
   s2p_sub = new s2p<bool,N>("s2p_sub",1);  // serial to parallel module
   s2p_sub->in(in);
   for(int i=0;i<N;i++)
     s2p_sub->out[i](sig_pa[i]);

    ...

    ifft_sub = new fft_ifft<N>("ifft_sub","IFFT"); //IFFT module
    for(int i=0;i<N;i++)
      {
	   ifft_sub->in_real[i](sig_real[i]);
	   ifft_sub->in_imag[i](sig_imag[i]);
	   ifft_sub->out_real[i](sig_out_real[i]);
	   ifft_sub->out_imag[i](sig_out_imag[i]);
      }

    ...

    p2s_r_sub = new p2s<double,N>("p2s_r_sub",1); //quadrature mixer module
q_mixer_tr_sub = new q_mixer_tr("q_mixer_tr_sub",mixer_fc,_amp,0.,0.,mixer_rate,false);
    q_mixer_tr_sub->i_in(sig_out_i);
    q_mixer_tr_sub->q_in(sig_out_q);
    q_mixer_tr_sub->out(out); /***/

  }
};

As we can see from the source code the modelling of an OFDM transmitter is quite easy when using the existing modules in the library.

Simulation example:
In this Section several simulation results are presented with different settings of the following parameters:

• constl_dim, defines the number of points in the QAM constellation
• ampl_se , defines the amplitude of the carrier for the transmitter
• ampl_re, defines the amplitude of the carrier for the receiver
• n_amp, maximal value of noise in channel
• atten, attenuation of channel

1. without noise and attenuation in channel:

constl_dim = 16,ampl_se=1.0,ampl_re=1.0,n_amp=0.0,atten=1.0

2. with gaussian noise ,without attenuation in channel:

constl_dim = 16,ampl_se=1.0,ampl_re=1.0,n_amp=90.0,atten=1.0

As we can see that because of the noise in the channel we got 5 error bits.

3. with gaussian noise and 50% attenuation in channel:

constl_dim = 16,ampl_se=1.0,ampl_re=1.0,n_amp=90.0,atten=0.5

Having both noise and attenuation in the channel the number of error bits became larger.

4. with gaussian noise and 50% attenuation, but with higher transmission power

constl_dim = 16,ampl_se=20,ampl_re=0.1,n_amp=90.0,atten=1.0

As presented in this figure we can reproduce the correct signals with higher transmission power. This is due to the parameter “ampl_se” set to 20, which means the signal from the transmitter is 20 times as strong as by previous simulations.

5. with gaussian noise and 50% attenuation, but slowing down transmission speed

constl_dim = 4,ampl_se=1.0,ampl_re=1.0,n_amp=0.0,atten=1.0

Or it is also possible to slow down the transmission speed to improve the BER. This is due  to the parameter “constl_dim” set to 4, which means now only 2 bits (which is 4 bits when setting “constl_dim” to 16) were encoded at one time.

For questions and suggestions regarding the building block library please contact