Simple data transmission over a distance where carrying data and clocking together is straight forward and the receiver doesn't have an issue with transmission speed until line impedance decays the signals. But what about fiber or microwave transmissions or another reason that the clocking is not a part of the scheme? How does the receiving end get the clocking lined up so the data is interpreted correctly.

In the past, phase lock loops have been the answer to adjust the phase of rising in data edges and flywheel the local oscillator to match and correct phase offset and align clock to data edges and then use a x2 multiplier within the loop and have a local clock to line up with the center of the data pulses. As transmission speeds get faster and data pulses narrow, it becomes harder to avoid jitter and data loss. The latest twist on PLL approach is to use an edge detector to create pulses for each rising or falling edge of the data stream and then search the spectral content for a fundamental frequency and that would be the basic clock frequency.

Now knowing what frequency to target, the PLL can go directly there and continually stay in sync. This improves the ability to sync and stay in sync with the data.

Fiber optic transmissions are basically pulses of light with inherent timing baked into the transmission. Detecting the spectral content of the bitstreams is a bit more dicy but it can be done with high speed photo-transistors.

With spectral density detectors added to PLL circuits, data synchronization can be realized for high speed transmissions.