Digital Signaling Principles—Part I

In Chapters 2 and 3 we learned that the first step in designing a communication system is to ensure adequate power transmission between locations. From engineering practice we developed some insight into how to optimize power transfer to obtain efficient communication links that can operate in the presence of noise and transmission impairments such as shadowing and multipath reception. In Chapter 4 we learned how frequency reuse enables the design of communication systems that serve multiple users over a growing geographical area. We are now ready to consider how information can be transferred across the links in such a communication system.

We postpone consideration of what the information might be until Chapter 7. At this point we wish to create a “pipe” that can be used in a transparent way to convey information from a source to a destination.

Because most communication systems being designed today carry information in digital form, this chapter is focused on digital signaling principles. We begin with a discussion of baseband digital communications and then extend our view to include carrier-based digital communications.

It is important to note in getting started that communication systems designed to carry information in digital form and communication systems designed to carry information in analog form are both analog systems. It is the information format that is digital or analog and not the signals transmitted over the wireless link. Conventionally, systems that carry information in digital format are known as “digital communication systems,” and we will use this terminology in this chapter.

Digital communication systems differ from analog systems in the design goal of the physical communication link. Analog links are designed with the object of ensuring that the user at the receiving end of the link is provided with a faithful replica of the information waveform. For digital links, the goal is to provide the user at the receiving end with a faithful replica of the information. Preserving the waveform is not important, unless doing so contributes to this goal.

The distinction between preserving waveform and preserving information will become clear as we investigate a simple baseband digital link. We will develop the basic indicator of performance quality, the bit error rate, and we will show how performance depends on signal design, as well as on other factors not entirely under the designer’s control. We will investigate optimal receiver design for the additive white Gaussian noise (AWGN) channel.

Once we have presented the baseband digital link, we will add modulation. Modulation provides a number of advantages, not the least of which is that it enables radio transmission. To implement a data link based on a modulated carrier, a link designer must choose both symbol and modulation formats. The choice of format has implications for spectral efficiency, implementation complexity, power efficiency, adjacent-channel interference, and flexibility that the designer must take into account. In most cases, a systems engineer works within the constraints of a link budget, cost and complexity limits, spectrum regulations, and packaging and power concerns to select the proper signaling format.

This chapter presents several forms of digital modulation in common use. Special emphasis is placed on visualizing the time waveforms and power spectra for each format so that operation and performance can be understood. The section ends with a discussion of the various performance measures used to compare and facilitate selection of a modulation format.

Finally, spread-spectrum signaling is introduced. Spread spectrum is a modulation technique that broadens the bandwidth of the transmitted signal in a manner unrelated to the information to be transmitted. Spread spectrum was originally developed to provide concealment and security, but was subsequently found to be very effective in making signals resilient in the presence of interference and frequency-selective fading. We will encounter spread-spectrum techniques again in Chapter 6 as they also provide an effective multiple-access technique.

Baseband Digital Signaling

A baseband signal is a signal whose energy is concentrated around DC. A digital baseband signal often contains binary information represented as some kind of pulse train. Information is typically presented to a transmitter as a bit stream encoded in baseband form prior to modulation.

In this section we introduce basic digital signaling architecture and techniques, performance analysis, and some typical improvement schemes.

Baseband Digital Communication Architecture

By definition a baseband digital communication system does not use modulation techniques to convey signals. As a consequence, it has the relatively simple architecture shown in Figure 5.1. Baseband digital communication systems or links may be used to convey signals between digital logic chips or between other wire-connected digital devices. The chief goal of the sending elements is to convert a stream of logical data into an analog waveform that can be sent along a tangible medium such as a pair of wires. The receiving elements examine the incoming analog waveforms, which may be noisy and distorted, and attempt to reconstruct the stream of logical data with a minimum of errors. The “logical data source” may produce signals that are inherently in a digital format, as in the case of a computer keyboard, or derived from an analog waveform that has been digitized. Techniques for digitizing waveforms, in particular speech waveforms, will be discussed in Chapter 7. The block in Figure 5.1 labeled “Encryption” has been included to suggest that the data from the source may be subject to digital processing prior to transmission.

A variety of digital processing steps may be included, such as data compression, encryption to ensure privacy and authenticity, and coding for error control. Error control coding is also discussed in Chapter 7.

Our focus in the current section is on the block labeled “Encoding,” whose job is to map the logical symbols into a series of analog pulses that make up a waveform known as a “line code.” At the receiving end of the link, the “Decoding” block reconstructs the stream of logical symbols. Next, digital processing steps such as error control decoding, decryption, and decompression are carried out before the data is delivered to its destination. We note in passing that each of the processing steps—digitization, compression, encryption, error control, and waveform generation—may be called “coding” in the literature. In the present chapter we confine the terminology to refer to the blocks labeled “Coding” and “Decoding” in Figure 5.1.

A line code  is a pulse train or series of “symbols” that represents digital data as an analog waveform. Each symbol has a fixed duration called the symbol period . For systems in which one symbol represents one bit of information, the line code is classified as a binary line code.

For a binary line code there are only two possible symbol waveforms that can be transmitted in each symbol period, one to represent a 1 and one to represent a 0. Some line codes utilize a set of more than two symbol waveforms that can represent multiple bits in a single symbol period.

There are many line codes in common use for representing digital data. Some typical forms are shown in Figure 5.2 where the bit sequence 10110010 is shown encoded four different ways. The upper left corner of the figure shows the familiar unipolar non-return-to-zero (NRZ) line code often used in transistor-transistor logic (TTL) circuits. The NRZ code gets its name from the fact that the voltage level does not return to zero during the symbol period. “Unipolar” refers to the fact that the two voltage levels used to represent the logical bits are zero and some nonzero (positive or negative) value. The upper right graph shows a polar NRZ line code. Polar codes use symmetric positive and negative voltage levels to encode 1s and 0s. The lower left plot shows a return-to-zero (RZ) code in which, as the name implies, the waveform returns to zero during the symbol period. This example shows an RZ code with a 50% duty cycle. In general, the duty cycle must be specified. The final plot, shown in the lower right corner of the figure, shows a Manchester code. This polar line code uses voltage transitions at the midsymbol period instead of voltage levels to carry information. For a symbol period of Ts , a 1 is represented as a transition from a negative to a positive voltage value at time Ts /2, and a 0 is represented as a transition from a positive to a negative voltage.

Each of these line codes has specific advantages and disadvantages. The unipolar NRZ line code is simple to implement and is easily extended to generate an “on-off keyed” modulated signal.

The polar line code requires two voltage reference levels at the transmitter, but it produces signals with a zero DC level, which can be important when the channel is transformer coupled. Also, the symmetry of the polar line code makes it suitable for differential-mode transmission, which aids noise immunity. The polar line code is easily extended to generate “binary phaseshift keyed” modulation, which we will encounter later in the chapter. Both the unipolar RZ and the Manchester code guarantee that there is at least one edge in each pulse regardless of the data pattern. This can be useful in helping the receiver synchronize to the bit pattern. Although the Manchester code is more complex, it offers the advantage of a zero DC level. Both the unipolar RZ and Manchester line codes occupy a wider bandwidth than do the two NRZ line codes. In the figure, all of the symbols have rectangular shapes, but other shapes are often used. We will be especially interested in pulse shaping aimed at reducing the bandwidth of the transmitted signal.

Multiple bits can be encoded into a single symbol for more bandwidth-efficient transmission. For example, in the 2B1Q scheme pairs of bits are encoded into a single four-level symbol.¹  The algorithm for encoding is demonstrated in the following example.

Next:  Baseband Pulse Detection

^1. 2B1Q stands for “two binary, one quaternary” symbol.

Introduction to Wireless Systems ?By Bruce A. Black, Philip S. DiPiazza, Bruce A. Ferguson, David R. Voltmer, Frederick C. Berry, Published Jun 7, 2011 by Prentice Hall, is reprinted with permission by Pearson Publishing.