Communication Theory And The Electrical Engineer
Having a strong understanding of communications theory is important for those who work in the field of electrical engineering. Basic theory has a tendency to find its way into just about every project. Keep real world applications in mind when looking at communication theory. This will help ensure that otherwise abstract topics will remain more concrete when learning about them.
Analysis of Different Types of Signals
Analyzing a deterministic signal is a concept that’s easy for most electrical engineering workers to understand. A deterministic signal is any signal that doesn’t have any uncertainty when it comes to its value in terms of any independent variables. Generally the variable engineering experts are concerned with is time. A rectangular pulse is an excellent example of a deterministic signal, and it’s certainly something that’s easy to visualize. Sinusoidal signals like sine and cosine waves are also considered deterministic.
Random signals lack these kinds of concrete metrics, and these pose a challenge for engineering teams designing systems to work with information. Models that consider random signals work with probabilistic concepts.
Consider a very simple digital communication system that has a binary symmetric channel with a code and decoder system. The channel the system uses introduces some sort of random bit errors. A code scheme is therefore put into play to combat these errors by using repetition. The decoder circuit then decides which bit was sent. It does this by way of a majority vote code rule. These kinds of systems are needed in order to prevent random signals, like those presented by atmospheric noise, from disrupting delicate communications patterns.
Amplitude modulation (AM) is one of the most basic ways of transmitting information over an electromagnetic wave. This technique varies a carrier wave in proportion to the waveform that’s being transmitted. The waveform can represent sounds that a loudspeaker reproduces, or the light intensity of pixels on a television display screen.
Generally, a useful modulated signal won’t consist of a single sine wave like most engineering textbooks illustrate. Fourier decomposition can be used, however, to express a signal as the sum of a number of sine waves at different phases, frequencies and amplitudes.
AM signals are comprised of a set of frequencies that are above the carrier frequency as well as those that are below. The above half are referred to as the upper sideband while those below are a lower sideband. There are forms of AM communication in which one sideband is suppressed in order to increase the signal strength in the other sideband.
While it might not be as familiar to electrical engineering experts as AM is, angle modulation is a class of analog modulation that alters phase instead of amplitude. Angle modulation techniques alter the angle of a sinusoidal carrier wave in order to send some sort of data. The modulating wave changes the angle of the sine-wave carrier.
Frequency modulation is by far the most common angle modulation technique. This technique revolves around a modulating signal, which causes a carrier frequency to vary along with it. These are variations are manipulated by the amplitude as well as the frequency of the original modulating wave.
Phase modulation is slightly more exotic. This technique allows engineering crews to encode data as variations in the instantaneous phase of a single carrier wave. This is widely used in the field of digital synthesizer music, and PM can supplement FM in synthesizer design. FM has a digital correspondence known as frequency-shift keying, which allows engineers to code digital data as an FM waveform. Phase-shift keying is a similar technology that uses PM techniques the same way.
Sampling in Signal Processing
Sampling refers to the reduction of a continuous signal into one that’s more discrete. A sound wave can be broken down into a sequence of samples represented by a discrete-time signal. Sampling can be done for nearly any metric, though time and quantization are the ones that most engineering texts concern themselves with.
There might be some data loss when a continuous function is broken down into a discrete sequence and then reproduced as a continuous function. The fidelity of the result is dependent on the sample-rate of the original sample.
The Nyquist-Shannon sampling theorem states that it’s necessary to sample a signal at twice its maximum frequency in order to preserve the full information in the signal. Physical limitations might ultimately prevent this from happening. This is where the field of quantization might come into play.
Quantization in Signal Processing
Quantization refers to the process of mapping input values to a smaller set that’s countable. Whenever a signal is converted from an analog wave to a digital sample, there’s usually going to be some sort of a loss. These errors are due to rounding or truncated data.
Mapping many quantities to comparatively few is inherently non-linear. Since the same output value is shared by a number of different input values, it’s generally not possible to recover the exact input value when given only an output value. The set of possible input values could be infinitely large. This makes them uncountable. Machines need some sort of countable finite array to work with.
Rounding quantization is generally seen as a good approximation when working with audio, video or images. The purpose of this process is to retain as much fidelity as possible while eliminating any unnecessary precision in order to keep things within practical limits.