The Challenge of "Almost" Synchronous Clocks

At the heart of PDH's biggest challenges is the concept of "plesiochronous" operation. Unlike a truly synchronous system where a single master clock governs every device, a PDH network consists of equipment with highly accurate but independent clocks. Even with the best technology, these clocks will inevitably have tiny frequency differences.

When a receiver with its own clock tries to read data from a transmitter running on a slightly different clock, a timing discrepancy occurs. To handle this, devices use elastic buffers. However, if one clock is consistently faster than the other, the buffer will either completely fill up (overflow) or completely empty (underflow).

Bit Slips: The Consequence of Overflow and Underflow

When a buffer overflows or underflows, the network is forced to either discard or repeat a block of data to reset the buffer. This event is known as a slip. A slip is the loss or repetition of an entire frame of data (e.g., 256 bits in an E1 stream) and is a major source of errors in PDH networks.

Impact of Slips on Different Services

• Voice (Uncompressed PCM): Slips are least noticeable here, typically perceived by the human ear as a minor "click" or "pop" in the audio. They generally do not affect the intelligibility of a conversation.
• Fax and Modem Data: These services are highly sensitive to timing. A single slip can corrupt a large portion of a transmission, almost always forcing a retransmission of a page (for faxes) or a block of data, leading to longer connection times.
• Digital Data (e.g., File Transfers): Slips are critical here, causing the loss of entire data blocks. This triggers higher-level protocols (like TCP) to retransmit the missing information, significantly reducing effective throughput.
• Compressed Video: A slip can be catastrophic for compressed video. Due to the inter-dependencies between video frames, a single lost block can cause severe, visible distortion (e.g., blocky artifacts, frozen picture) that can persist for several seconds until a new keyframe is received.

Managing Slips: The Hypothetical Reference Connection (HRX)

Since eliminating slips entirely is impossible in a plesiochronous network, international standards bodies like the ITU-T defined acceptable performance levels. They did this using a model called the Hypothetical Reference Connection (HRX).

The HRX is a standardized, theoretical end-to-end international connection with a total length of 27,500 km. Standards like ITU-T G.822 specify the maximum acceptable number of slips per day for a connection of this length. This model provides a benchmark against which operators can measure the performance of their real-world networks. For example, a common requirement might be no more than one slip in 70 days for the entire connection, with stricter limits on individual nodes within the chain.

The Path to a Solution: Synchronization

The accumulation of timing errors and the resulting slips across long chains of PDH equipment was a major limitation. It highlighted the need for a different approach. Instead of compensating for clock differences at every step, the next generation of technology, SDH/SONET, was designed to be fully synchronous.

In a synchronous network, all devices are timed by a single, highly accurate master clock. This network-wide timing eliminates the root cause of plesiochronous slips, enabling much higher levels of performance and reliability.