Transmission Basics  |  Components  |  Applying Fiber Optic Technology  |  System Performance  

Digital Modulation

Figure 1 - Block Diagram of a Basic Digital System

Basic Digital System
Digital transmission requires converting the analog inputs to digital pulse-code modulation (PCM) data signals. The PCM data is "line coded" such as scrambled NRZ to simplify time synchronizing, critical to digital television (DTV) transmission, as well as providing error monitoring and economic use of the available bandwidth. This line-coded PCM signal modulates the light source. This conversion use precision analog-to-digital circuits. Figure 1 illustrates a basic diagram of a digital transmission system. Some variants of PCM are listed below. The first three describe analog-to-digital modulation, and the last two are strictly digital modulation schemes. Pulse-amplitude Modulation (PAM): Information is encoded by a stream of pulses with discrete amplitudes. Delta Modulation (DM): Pulses are sent at a constant rate with duration determined by the first derivative of the input signal. Adaptive Delta Modulation (ADM): Similar to DM with the ability to adjust the slope of the tracking signal. Phase-shift Keying (PSK): Information is sent over a constant carrier frequency. The phase of the carrier is shifted between two levels as determined by the digital bit to be sent. Differential-phase Shift Keying (DPSK): A variant of PSK that allows for more straightforward decoding. Pulse-code modulation, as the name implies, involves the assignment of a sequence of pulses (or a code) to represent a portion of a signal. Two representations of amplitude include the actual analog voltage and a 4-bit binary digital voltage. Four bits is defined as the resolution or accuracy of the code, allowing up to 15 voltage increments to be represented. These voltages and 4-bit codes are listed in Table 1.
Table 1 - Amplitude Coding
Increment Binary Code Voltage
(Voltage Increment = 0.066 Volts)
0 000 0 Volts
--- --- ---
--- --- ---
--- --- ---
7 0111 0.46 Volts
8 1000 0.53 Volts
--- --- ---
--- --- ---
--- --- ---
15 1111 1.0 Volts

The pulse code uses four separate pieces of information to convey a single data stream. 0.53 Volts has become 1000. More specifically, the analog signal has been sampled and a series of voltage measurements have been converted to the code shown Table 1. Notice that the voltage does not break cleanly at 0.50 Volts. In reality, there is no 4-bit digital code for 0.50 Volts, only 0.46 Volts and 0.53 Volts. (A longer bitword would be required to accurately code 0.50 Volts.) The inability of the 4-bit word to accurately describe 0.5 Volts is called quantization. Quantizing is dynamically related to the slope of the sampled signal. In quiet, unchanging portions of the video picture, the effects of quantizing are not apparent. In fact, the quantizing effects actually mask random noise components which are less than 1 least significant bit. Quantizing and the resultant pulse codes that are generated "freeze" the signal in a form that is very tolerant of subsequent transmission or processing.
Figure 2 — Time-division Multiplexing

Time-division Multiplexing
Once the analog information has been put into a digital form, the digital channels are time-division multiplexed (TDM) and sent to the laser transmitter. The digital signal is converted into light pulses; the laser is on for a "1" and off for a "0." Time-division multiplexing is used by digital systems to either combine multiple video signals on to one fiber or to create subchannels for digitized audio and/or data signals. TDM allows signals to be added to or removed from a system without system degradation Figure 2 illustrates the process of TDM. At the receiving end, the light pulses are converted back into electrical pulses. The pulses are time-division demultiplexed (TDD) and sent through a network digital-to-analog (D/A). This converts the information back into a baseband video signal. As in all digital modulation systems, sample rate and accuracy affect the end-to-end signal performance of the system. However, unlike analog systems, the performance is not further affected by transmission distance, light source noise, or distortion. Digital modulation does not require a linear light source, allowing the light source of a digital system to have a wide range of non-critical operating parameters. Digital modulation is noise immune, so system performance will not degrade over very long ranges, eliminating the need for repeaters. Digital signals can also tolerate losses and reflections from connector, splices, and optical devices such as splitters and wavelength-division multiplexers (WDM). The accurate transmission of a digitally modulated signal ultimately relies on the optical receiver's ability to detect the transition between a "1" and a "0" in the data stream, a function called a decision circuit. To assure this ability, the data stream is "conditioned," or line encoded, using one of several methods known collectively as scrambling. Scrambling rearranges the "1's" and "0's" of the data stream into a predetermined manner, depending on the structure of the input data. Line coding also ensures that the incoming data to the receiver clock recovery subsystem is balanced in bit sequence to allow the phase shifts, caused by uneven data patterns, to be avoided.