Synchronization in a Wideband Optical Data Transmission System

An optical synchronization technique is described for the demultiplexer of a wideband optical data transmission system. In this system, the closely spaced optical pulses of interleaved PCM channels are “space sorted” by a coincidence detection technique using an optical reference waveform generated in the receiver. For synchronization of the reference waveform with a spatial waveform produced by the received laser beam in the demultiplexer, a number of pulses on these two optical waveforms are coded with a pseudorandom sequence. Acquisition of the coded waveforms for synchronization is indicated in an optical matched filter. For automatic tracking, a synchronization error-control circuit is added. This communication suggests an optical correlation method for synchronization in an optical data transmission system in which many pulse-code modulated (PCM) channels are time-multiplexed [ 1 31. In the particular type of system discussed here, the sequence of data pulses transmitted during each time frame is derived from one pulse of a mode-locked laser[4]. Each time frame contains pulses used for the acquisition and error control of synchronization as well as the pulses that are single data bits of the interleaved PCM channels. Optical seriesto-parallel conversion in the demultiplexer (see Fig. 1) transforms the received serial data stream into a succession of “spatial waveforms,” but only one of the spatial waveforms, H, , contains a sequence of pulses that is identical to that in the time frame of a received laser beam. With the aid of an optical reference waveform, which is generated in the demultiplexer and is in synchronism with H, , the PCM channels are “space sorted” by coincidence detection[5,6] and the data bits in H , are routed to the correct channel receivers. The problem of synchronization becomes especially important when high channel capacity is desired, since the time-multiplexed optical pulses can be spaced so closely that they cannot be separated by electro-optical techniques. The use of an optical correlation method is one way to solve the problem. In the method proposed here a number of pulses in each frame of the transmitted and reference waveforms are coded with identical pseude random sequences [7]. Synchronization is indicated by 408 the autocorrelation maximum, which occurs when the two coded waveforms are in time coincidence. The autocorrelation function of the pseudo-random sequence has the general form where the a’s are either 1 or 0, N is the length of the pseudo-random sequence and n is the number of 1’s in the sequence. On the coded optical waveforms following the pseudo-random sequence, the 1’s are represented by optical pulses and the 0’s by their omission. The autocorrelation enhancement when the two waveforms are in time coincidence (k = 0), is equal to the number of optical pulses. The subsidiary maxima, which can occur when the two waveforms are offset by k = 1 to N 1 bit positions, can never exceed one. We will assume that the correlation method is used in an optical system operating with a continuously modelocked Nd-YAG laser having a pulse repetition frequency f = 200 MHz and a pulse width 7 = 25 ps. With a separation of 25 ps, the number of time-multiplexed pulses can be 100. Through the use of pulse delay and modulation, the serial stream of data transmitted during each time frame can have the form 1 0 1 0 0 1 0 0 0 1 0 0 0 0 0 1 1 1 d , d ; . * d s z , (2) synchronization synchroPCM , \ sequence nization channels error