Error conditions

Principal alarms

T1 and PCM-30 alarms and error conditions

The principal alarms defined by both Tl and PCM-30 are the red alarm, generated by the receiving equipment to indicate that it has lost frame alignment, and the yellow alarm, which is returned to the transmitting terminal to indicate that the receiving terminal has lost frame alignment. Normally, the terminal will use the receiver's red alarm to request that a yellow alarm be transmitted. The name of these alarms simply comes from the colour of the lights on the original equipment.

Loss of frame alignment is detected simply by monitoring the frame alignment signal. PCM-30 differentiates between loss of frame alignment, being a failure to synchronise on the frame align­ment signal FAS (0011011), and loss of multiframe alignment, which is caused by a failure to synchronise on the multiframe alignment signal (0000) contained in bits 1 to 4 of time slot 16 of frame 0. The former is the red alarm, the latter is the multiframe red alarm. Table 21.4 illustrates how these alarm conditions are trans­mitted. Tl multiplexers can also signal a blue alarm. This is a continuous ones pattern across all 24 channels (the F-bits, however, remain unchanged) to indicate an upstream failure.

The most basic impairment that Tl orEl equipment can suffer from are bipolar violations (BPVs). A bipolar violation is a violation of bipolar coding in which two pulses occur consecutively with the same polarity. As will be discussed later in this chapter, Tl and El signals are encoded with a system that inverts the polarity of alternate one bits so that two pulses of the same polarity will not occur in a row. On metallic circuits, a bipolar violation will occur if a zero is changed to a one, or if a one is changed to a zero. Since bipolar violations occur one for one with bit errors, the BPV rate (the ratio of BPVs to correct bits) corresponds to the bit error rate.

CRC errors are a second possible Tl/El error condition. The cyclic redundancy check error measurement is an alternative to frame error measurement, and is available to Tl circuits that employ the Extended Superframe (ESF) format and to El circuits that employ the CRC-4 multiframing.

CRC-n (n = 4 with PCM-30, n = 6 with Tl ESF) is an error checking method that uses an n-bit code to represent an entire multiframe of data bits. The n-bit code is arrived at by applying a complex mathematical function to each group of 24 (ESF) or 8 (PCM-30) frames of data. The result of this calculation, the n-bit code, is then transmitted in the CRC framing bit positions of the following frame.

At the other end of the circuit, the same mathematical function is performed on the same first group of frames. This newly calculated n-bit code is compared with the code that was calculated by the transmitting equipment. Any discrepancies between the two codes are counted as CRC errors. CRC errors are by far the best in service performance measurements because the CRC scheme allows detec­tion of errors on all of the data bits within each group of frames, with an accuracy of about 98 percent.

Slips are yet another impairment. Slips are the most common impairments caused by frequency deviations and timing problems. A slip is the insertion or deletion of data bits into or from the data stream. It is the direct result of equipment buffer overflow or underflow, resulting from improperly timed network equipment. Digital equipment uses input buffers of finite length to accommo­date the momentary frequency fluctuations that can occur between the receiver and transmitter of a digital network node. These fre­quency differences become persistent when connecting to (or se­lecting) the wrong clocking source, causing buffers to overflow or underflow and then reset themselves. It is the buffer resetting that results in the addition or deletion of data bits from the bit stream.

Based on the source of the slips and their effects on the network, all slips can be placed in one of two categories, controlled slips or uncontrolled slips. Controlled slips are bit additions or deletions that do not disrupt frame synchronisation. Uncontrolled slips are bit additions or deletions that cause both framing and data to be dis­placed. This framing and data displacement results in a loss of frame synchronisation, effectively taking the circuit down momentarily.

Jitter, the cyclic offset of bits from their expected positions in time, is one of the most ominous of all Tl/El impairments. Jitter can be intermittent and data dependent, which makes it difficult to isolate. Jitter occurs little by little, cumulatively over many bits, and can ultimately cause the missampling of the pulses resulting in bipolar violations and bit error conditions.

The most common cause of jitter is the network equipment itself. Jitter is inherent to the clock recovery timing used in transmission, and is typically added to the pulses at every regeneration point within the network. As long as each network component adds only a very small amount of jitter, the circuit will be unaffected. Prob­lems arise when a failed or failing network component adds signi­ficant amounts of jitter. Less typically, jitter can come from crosstalk, electrical noise, and other types of interference.

Wander is an impairment very similar to jitter, and is defined as jitter occurring at a frequency of less than l0 Hz. Wander, like jitter, is a back and forth (cyclic) displacement of bits from their expected positions. But, because wander occurs at such low frequencies, its cause within the network and effects on the network are very different from those of jitter. Wander is most often caused by instabilities in a master timing source, or by nocturnal cooling (cooling as the sun goes down). The end result is usually slips.

Дата добавления: 2015-03-31