Digital Signal Processing - Report Example

Running Head: Digital Signal Processing Digital Signal Processing [Writer] [Date] Digital Signal Processing Digital Signal Processing At the time when data transmitted from one place to other in signal form, it may be possible that it become complex. For this reason, it is quite essential to maintain this data in a format that is easy to transmit and store in a simple manner as well. This is obviously helpful to keep the important information secure (N.B. Jones, J. D. McK, 1990). What is a Digital Signal? If we go to the past, it would become obviously clear to us that historically signal processing comes from electrical engineering where electrical signals were transmitted through a telephone line or a wire otherwise by a radio wave. The term ‘Digital’ in digital signal processing straightforwardly refers to ‘numbers’ and in French it is called as ‘numerique’. Therefore, a digital signal holds data in form of streams of numbers mostly in the binary format and the data in this form is processed by numerical calculations. Digital signal processing has been a part of most electronic mechanisms across the globe. Nowadays, digital signal processing technique is widely used for making digitally significant systems. Signal Processing: Digital signal processing is basically a way to process data (numbers, symbols etc) in the form of signals. Digital signal processing is a sub field of Signal Processing where its other sub field is analog signal processing. Digital signal processing is further divided into subcategories i.e. Sensor array processing Audio and speech signal processing Radar signal processing Biomedical signal processing Seismic data processing etc. Procedure: In this process of digital signal processing, first signal is converted from analog to digital form which is usually done with the help of an analog to digital converter. Usually, in digital processing analog signals are converted into digital signals. Certainly digital form of a signal makes it easy to detect the errors and correct them as well and thus comes with less noise as compare to the analog signals. In the current era, several latest systems and devices have been introduced for signal processing such as FPGA or Field programmable gate array, microprocessors, stream processors and microprocessors. Although another scheme used for this purpose comprise DSP algorithms. Digital Signal Processing domains There are certain digital signal processing domains that are helpful to process data or the given signal. The most important and popular signal domains are: Time domain that is probably important for one dimensional signal Spatial domain for multidimensional signals Frequency domain Wavelet domain Autocorrelation domains Choosing between the DSP Domains: The selection among these available signal domains purely depends on the required attributes or characteristics of the signal. A measuring device producing a sequence of samples creates a time or even a spatial domain. But for frequency domain, discrete Fourier transformation is practical and realistic. Signal Sampling With the passage of time, the principally use of computers has more signified the importance of digital signal processing. And for this reason, a computer must have to be digitized with an analog to digital converter to employ an analog signal. This conversion is somewhat called as signal sampling. Signal sampling is usually performed in two phase, one is discretization and the other one is quantization. In first phase of signal sapling i.e. in discretization, the available signal space is divided into correspondent classes which then in second phase i.e. quantization are replaced by the signals belonging to a relevant equivalence class Digital Signal Processing Implementation When talking about the implementation of digital signal processing some particular microprocessors are very important in this regard for instance: The DSP56000 The SHARC The TMS320 The listed microprocessors make use of fixed point arithmetic to process data. Although FPGAs; are popular for fast applications. On the other hand, for slow applications, a conventional slower processor may be obliging. Digital Filers: In a number of devices and systems (based on signal processing and information technology) Digital Filters are playing a fundamental part (Marek, 2005). What is a Digital Filter? Definition: Before we continue to discuss the design of a digital filter, let us first know what digital filters are? A Digital Filter is a device or a system that carries out several mathematical operations on a signal (sampled and time discrete). The basic purpose of a digital signal is to decrease or in some cases, to augment some features of a signal. A digital filter system generally composed of an analog-to-digital converter a microprocessor (usually a specialized digital signal processor), and one digital-to-analog converter which performs signal converting functions. Software executing on the microprocessor can implement the digital filter by conducting the necessary mathematical operations on the numbers executing from the ADC. In few high performance applications, an FPGA or ASIC is widely used despite of a general purpose microprocessor. In addition, a digital filter can also work with the analog signals by first digitizing them and then represent them in a sequence of numbers. After that, mathematical operations are applied to the signals and finally it is reformed as a new analog signal. You will find there in digital filter system design: An analog to digital converter (for sampling the input signal) A digital to analog converter A microprocessor In contrast to the ordinary analog filters, digital filters are much efficient and practical and work in several conditions where analog filters impede. Although they are quite expensive as compare to the analog filters as they involve complex functionality. Mostly digital filters face latency unlike the analog filters where latency means the time difference between the input signal and its response. Analog filters offer less latency sometimes small enough to be neglect able Digital filters have now become the important component of everyday electronics for instance they are used in cell phones, radios, and stereo receivers etc. Digital filters can be more expensive than an equivalent analog filter because of their increased complexity, but they develop practical many designs that are unpractical or impossible as analog filters. Since digital filters utilizes a sampling process and discrete-time processing, they usually experience latency (known as the difference in time between the input and the output), that is almost unrelated in analog filters. Types of Digital Filters: In a number of digital filters, Fast Fourier transformation is employed which is a mathematical algorithm that extorts the signal spectrum to manipulate it earlier than converting it back into a time series signal. Furthermore, state-space model is another type of digital filters. Digital Filters Design: Designing a digital filter required some definite specification that guide about the preferred performance of the filter. This specification should identify the design measures. Some discrete time properties are too significant here (Roland, 1991). Now taking into consideration a specific digital filter we will see a number of factors associated with a digital filter such as it’s: Sampling rate Ripple Attenuation Cutoff Frequency etc Sampling and Bandwidth: Obviously the larger the sampling rate of a digital filter, the more bandwidth it will produce. For instance a digital filter with sampling rate 16k Hz is capable of producing bandwidth greater than 7 k Hz (Steve, 2002). So, to obtain better results, it is suggested to use a digital filter with higher sampling rate. Cutoff frequency: Another characteristic of a digital filter is Cutoff frequency. Cutoff frequency of a digital filter is a frequency which develops an edge between a pass band and a stop band. In the majority cases, this frequency provides a site where a transition band and pass band usually meets e.g. cutoff frequency of 7k and alike. High pass: A digital filter with high pass enables the high frequency signals to easily pass from the source, whereas low frequency signals difficultly pass through it. A high pass digital filter tends to pass high frequencies while reducing the amplitude of these frequencies, this is known as attenuation which is taken as a design constraint for the filter. In this way maximum stop band attenuation is produced. These types of digital filters are called as low cut filters. In addition, a digital filter with data frequency up to 800 Hz along with the ripple less than 0.05 is able to produce maximum stop band attenuation. The stop band attenuation of a digital filter is basically the least attenuation within the stop band. All these factors make the digital filter efficient enough to work according to a certain condition. Now we will see the benefits of a high pass, 16 kHz sampling digital filter in the subsequent section: Applications of High pass Digital Filters: The popular high pass digital filters include Rumble filters that take away the noise. High pass filters are used in digital image processing for spatial frequency transformation. Furthermore, high pass digital filters are important for AC coupling as well. References Steve Winder. (2002). Analog and digital filter design EDN series for design engineers Referex Engineering. Newnes Marek Kurzynski. (2005). Computer recognition systems: proceedings of the 4th International Conference on Computer Recognition Systems, CORES'05 Volume 30 of Advances in soft computing. Springer N. B. Jones, J. D. McK. Watson, (1990). Digital signal processing: principles, devices, and applications Volume 42 of IEE control engineering series. IET Roland Priemer. (1991). Introductory signal processing. World Scientific Read More