DSP

In this experiment, we performed auto-correlation and cross-correlation. Apart from this we also used Carl Pearson Correlation Coefficient calculation using the provided formula and using FFT for the same. Correlation is used for checking the similarity between two signals, and hence is an important tool. DSP is quite efficient in giving us the computational results.

In this experiment, we performed linear convolution, circular convolution and linear convolution by circular convolution and both using FFT. In circular convolution, we observed that the output signal was just the addition of elements of the linear convolution output. We can see that the linear convolution signal was overlapped to give circular convolution output. This is aliasing. Therefore, we made the following conclusions from the above experiment: ● Length of linear convolution output signal= L+M-1 ● Length of circular convolution=Max(L,M) ● Circular convolution gives aliased output

In this experiment we want to implement Linear filtering Tradition methods of time domain based convolution require the entire input sequence to be present before commutation. This can pose a problem in real time applications where the data coming is long and continuous. Overlap save method and overlap add method divide the input sequence into smaller blocks. The filtering of individual blocks which are smaller in length is done by DFT and IDFT and then fitted one after the other. The overlapped portion is discarded in OSM whereas it is added in OAM to get the required output sequence. This resultant is the same as that if the input had been processed via time domain convolution.

Fourier transform is an essential part of signal processing , since it transforms a signal from the time domain into the Frequency domain. When we have a discrete signal we need to use Discrete Fourier transform is DFT FFT is an algorithm to compute DFT of any input FFT is a proven faster method as compared to the DFT algorithm. Some of the similar complexities re-appeared in this code too (example - integration of complex numbers). Unlike the DFT, i could identify patterns in the butterfly diagram taught in the class. Formulating this pattern into a code format and making this code generic(for 4-point and 8-point) and efficient was the real highlight of this code. Though a generic N-point FFT code, still remains an ambitious task.

In this experiment we get a to implement flash programming to implement a LED counter on the DSP board of TI, we understood the basics of flash programming and proper use of JTAG. In the given experiment, we utilized the on board LEDs to show the binary counting action happening, from 0-15 in binary. Once the flash programming was done, even after the power was removed and external power source was used, the DSP board still contained and ran the same code, hence we saw that if we want the same code to run on the Board, without interruptions then Flash programming is our go-to choice.

In a digital signal processing application the sensors are crucial part of the system. sensors are the means by which the processor aquires the data from outside world. generally sensors give output voltage in a fixed range depending on the sensor parameters. In this practical we used different types of sensors and interfaced them with the DSP by using an a/d converter. this practical enabled us to aquire real world data by using various kinds of sensors. We used an LDR which detects light, Optocoupler which is used to detect block in a certain pathway of light, Ultrasonic sensor and detector used to measure distance, Hall effect sensor used to measure magnetic fields, Audio sensor(microphone) used to capture surrounding audio, Temperature sensor to get the ambient temperature. Using all these sensors we can build real time standalone systems