IV. Analysis and Conclusion

The program was run several times in order to verify the performance of the program. The input signals were generated by Phillips PM5193 Programmable Synthesizer/Function Generator, and the comparison between the Input Signals and the Output Signals were done with Phillips PM3365 Digital Storage Oscilloscope. The input signals were generated in various forms, sine, triangular, rectangular wave, and pulse code with 1 V p-p and 2.9V dc at 100Hz. Following graphs are the results from various input signals.

As above, the companded signal has distortions in all sequence, especially around the peak amplitude and it can be concluded that these distortions, quantization errors, are occurred due to the loss of the precision over the process. These errors are unavoidable due to the limitation of the processor (32-bit arithmetic) unless the sizes of the buffers or memories are increased in bit-wise.
In terms of reducing SNR (Signal-Noise-Ratio) less than 40dB, it was possible to prove that compander reduced it because of the overall amplitude of companded signal, which was less than the original transmitted signal.
From the above results, it is possible to state that the compander reduces quantization error and overall signal-noise-ratio over its process, however, it is hard to reduce the quantization error close to zero due to the limitation of the hardware. There are several ways to reduce more quantization errors, such as expanding the buffer size of the hardware, along with the increment of the bit-rate of the telecommunication standards, but moreover, the best way would be to find more efficient companding rule than those of existing ones.

Previous Page Next Page