## 1 Mention The Different Types of Filters Based on Impulse Response

Filters are normally based on either impulse or frequency response. Do not, therefore, confuse these two. Kindly stick to those based on impulse response and answer the question confidently. You can also expound a little on them.

**Sample Answer**

There are two filters based on impulse response, i.e., IIR and FIR filters. IIR filters are normally recursive. Their present output sample relies on the current input, output, and past inputs samples. On the other hand, FIR filters are non-recursive, meaning that their present output samples are dependent on the present input and previous input samples.

## 10 DSP(Digital Signal Processing) interview questions and answers | DSP Questionnaire

This page describes DSP questionnaire written by specialists in DSP embedded domain. This top 10 **DSP interview questions and answers** will help interviewee pass the job interview for DSP Handware or software job position with ease.

This set of questions are very useful for passing interview for DSP engineer job position for fresher as well as experienced category. These questions are very useful as DSP viva questions also.

Question -1: Explain DSP basic architecture. Mention how to select DSP architecture suitable for algorithm under development. Answer -1: Refer DSP Architecture basics➤.

Question -2: Explain Harvard and Von Neumann architecture and mention difference between them. Answer -2: Refer Harvard Vs von Neumann➤.

Question -3: Explain difference between convolution and correlation. Answer -3: Refer Convolution➤, Correlation➤

Question -4: Mention FIR filter and IIR filter basics and differentiate FIR and IIR filter concepts. Answer -4: Both are digital type of filters. FIR stands for Finite Impulse Response and IIR stands for Infinite Impulse Response. In FIR filter, no feedback is used while in IIR filter feedback is provided at input. Hence FIR filter response is finite while IIR filter response is infinite. Refer FIR vs IIR filter➤ and Refer FIR MATLAB Code➤.

Question -5: Explain Interpolation and decimation and their applications in Digital Signal Processing. Answer -5: Interpolation increases data rate, decimation decreases data rate. Interpolation adds samples in between, Decimation removes samples from within. Refer Interpolation➤ and Decimation➤.

Question:Explain 8085 Microprocessor Architecture. Answer: Refer 8085 Architecture➤

Question -6: What is FFT? How it differs from DFT? Explain simple FFT implementation using MATLAB code. Answer -6: FFT stands for Fast Fourier Transform. DFT stands for Discrete Fourier Transform. FFT performs faster compare to DFT. Refer FFT DFT basics➤ and 16 point FFT decimation in frequency MATLAB Code➤

Question -7: What is the difference between microprocessor and DSP? Answer -7: Microprocessors are usually generic processor used in desktops, laptops etc. DSP are special processor designed for specific complex applications such as used in mobile, smart phone, tablets etc. Most of the DSPs will have special instruction set which takes care of complex signal processing in very less time and with the use of very small memory. Popular DSP vendors are Texas (TMS320), Qualcomm, LSI, Tensilica etc. Refer What is inside DSP➤ and DSP Chip Vendors➤.

Question -8: Often it is asked to convert floating point number to fixed point format. Answer -8: Refer Floating to fixed point conversion➤.

Question -9: Explain handling of interrupts in DSP? Answer -9: This is very useful when porting protocol stack on DSP and FPGA and when it is required to interface one layer(e.g. MAC) and the other layer (PHY). Synchronization between two layers have been done using interrupt concept.

**Question -10**: Explain how DSP chip can be interfaced with FPGA chip? **Answer -10**: One need to explain interfacing concepts used for data communication between FPGA and DSP chips.

In addition to DSP interview questions and answers, one can also refer architecture of any one of the DSP.

## What Do You Understand by Sectional Convolution?

This is a common concept in digital signal processing that you should expect in your upcoming interview. You must have probably done it at one point and therefore understand what it entails. Ensure that you illustrate your answer to the interviewer to enhance understanding.

**Sample Answer**

Sectional convolution is dividing a data sequence into smaller sections that are then independently processed and later controlled to get an output. This normally happens when the data sequence is of a longer duration making the output sequence quite difficult to obtain due to limited computer memory.