The Delay Locked Loop

After initial acquisition, the spread spectrum receiver must maintain synchronisation by tracking changes in the transmitter's PN code clock. The circuitry required is known as a tracking loop, as it tracks the transmitter's code clock frequency variations. Without a tracking loop synchronisation will be lost as the transmitter and receiver PN code clocks will tend to drift apart.

In a delay locked loop two identical pseudo-random or PN despreading codes are delayed with respect to each other. Each PN code is used in separate correlators (early and late) to despread (correlate) the received direct sequence signal. The result of correlation between an incoming direct sequence signal and the receiver PN code is a triangular function two chips (code bits) wide. Assuming synchronisation two correlated signals (each with a triangular correlation waveform) are produced with their correlation peaks separated by the delay between the early and late receiver PN codes. If the two correlation signals are summed in a difference amplifier and filtered, then a composite correlation function is produced. This composite correlation function has a linear region between its maximum and minimum values.

If this composite correlation function is used to control the receiver's code clock frequency (for example by driving a voltage controlled oscillator) then the receiver will track the transmitter's code clock at a point halfway between the maximum and minimum values of the composite correlation function.

An optimum solution is to have a third on-time (punctual) PN sequence correlator channel for signal recovery, with early and late correlators simply providing tracking to keep the on-time channel in the middle of the correlation window. Such an approach provides an optimally correlated (despread) output signal for subsequent data demodulation.

Copyright James A. Vincent, 1993