BER monitoring circuit

Abstract

In a BER monitoring circuit, error cycles of input data are detected by a parity check portion and an error cycle detecting portion, a maximum (average/median) value is detected from among the error cycles by an error cycle memory and an error cycle maximum (average/median) value retrieving portion. The value is converted into a corresponding estimated error rate by a Te-BER conversion table and an alarm is generated by an SF/SD detecting -portion when the estimated error rate exceeds an alarm detecting threshold. Thereafter, the alarm is released when the estimated error rate assumes equal to or less than an alarm releasing threshold. Also, an error-free detecting portion is activated when an alarm is generated and releases the alarm when a time period for which the error cycles stay flat exceeds a cycle corresponding to the alarm releasing threshold.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other objects and advantages of the invention will be apparent upon consideration of the following detailed description, taken in conjunction with the accompanying drawings, in which the reference numerals refer to like parts throughout and in which:

FIG. 1 is a block diagram showing an embodiment [1] of a BER monitoring circuit according to the present invention;

FIG. 2 is a diagram showing an error cycle T_e-BER conversion table used in a BER monitoring circuit according to the present invention;

FIG. 3 is a graph showing the T_e-BER conversion table shown in FIG. 2;

FIG. 4 is a block diagram showing an embodiment [2] of a BER monitoring circuit according to the present invention;

FIG. 5 is a block diagram showing an embodiment [3] of a BER monitoring circuit according to the present invention;

FIG. 6 is a block diagram showing an embodiment [4] of a BER monitoring circuit according to the present invention;

FIG. 7 is a block diagram showing an example of a prior art BER monitoring circuit; and

FIGS. 8A and 8B are diagrams showing a signal frame format of an OTU used by a parity check portion.

DESCRIPTION OF THE EMBODIMENTS

Hereinafter, embodiments of a BER monitoring circuit and method according to the present invention will be described referring to the attached figures.

Embodiment [1]: FIG. 1

FIG. 1 shows an embodiment [1] of a BER monitoring circuit and method according to the present invention. In this embodiment, input data is provided to a parity check portion 1, and an error pulse P_E outputted from the parity check portion 1 is transmitted to an error cycle detecting portion 2. A cycle monitoring reference clock CLK is provided to the error cycle detecting portion 2. Error cycles (number of frames) T_e that are output signals of the error cycle detecting portion 2 are provided to an error cycle memory 3.

The error cycle memory 3 stores a plurality of the error cycles T_e from the error cycle detecting portion 2, and the maximum error cycle T_emax thereof is detected by an error cycle maximum value retrieving portion 4 connected to the error cycle memory 3. The maximum error cycle T_emax is transmitted to a T_e-BER conversion table 5, converted to an estimated error rate BER_max corresponding thereto, and outputted to an SF/SD detecting portion 6.

The SF/SD detecting portion 6 is preliminarily provided with an alarm detecting threshold Th_d and an alarm releasing threshold Th_r (Th_r<Th_d), and generates and releases the SF/SD alarm ALM under given conditions.

The alarm ALM is also provided to an error-free detecting portion 7. The error-free detecting portion 7 further receives the error cycles T_e from the error cycle detecting portion 2 and the cycle monitoring reference clock CLK, and commonly provides an error-free signal E_f to the error cycle memory 3 and the SF/SD detecting portion 6.

In operation of this BER monitoring circuit, when the input data is provided to the parity check portion 1, the parity check portion 1, in the same way as the prior art example shown in FIG. 7, detects the parity error of the input data based on a parity check result with respect to the BIP-8 calculation field in the signal format of FIGS. 8A and 8B, and provides an error pulse P_E to the error cycle detecting portion 2. It is to be noted that when the BIP-8 and the like are processed in parallel, the output of the parity check portion 1 may be made a total of the number of error counts and a latch timing.

In the error cycle detecting portion 2, the cycle monitoring reference clock CLK is used for counting the number of reference clocks from the last inputted error pulse to the present error pulse. The counted value is outputted to the error cycle memory 3 as the error cycle (number of frames) T_e of a binary number. At the same time, the error cycle detecting portion 2 outputs a write request pulse W_req indicated by a dotted line to the error cycle memory 3. It is to be noted that the cycle monitoring reference clock CLK is synchronized with a parity calculation cycle.

The error cycle memory 3 sequentially stores the error cycle T_e upon receiving the write request W_req from the error cycle detecting portion 2. In this case, the error cycle memory 3 has a storage capacity for e.g. 80 times in total.

The error cycle maximum value retrieving portion 4 provides a read request R_req indicated by a dotted line to the error cycle memory 3 at a constant frequency, for example, to read all of the data within the memory 3, and then selects the maximum value from among all of the error cycles T_e read. The maximum value read is transmitted to the conversion table 5 in the form of a binary number T_emax.

The conversion table 5 outputs a BER estimated value BER_max based on and corresponding to the maximum value T_emax of the error cycle provided from the error cycle maximum value retrieving portion 4.

FIG. 2 shows an example of the conversion table 5 mentioned above. For each of the data of the error cycle maximum value T_emax, the BER on the left side thereof indicates the estimated error rate value. For example, the maximum value T_emax=60,306.750 shown by hatching corresponds to BER_max1=2×10 ⁻⁹. As will be described later, the reason why the maximum values T_emax upon alarm detection and upon alarm release are different is that a hysteresis is provided therebetween. In this example, the estimated error rate upon alarm release (BER_max2 for BER_max1) is set to one tenth of the estimated error rate upon error detection (e.g. BER_max1).

A calculation example for having the BER from the error cycle will now be described as follows:

The probability that a parity error of “r” bits within a frame occurs is given by the following equation:

P(r)=_NC_r×((1−q)^(N−r)×q^(r))   Eq. (1)

where,

N: Number of bits to be calculated for 1 parity bit

q: Line error rate

When a probability including the event probability of a plurality of bits is obtained, a probability of detecting a parity error within 1 parity frame assumes ΣP(r). It is to be noted that the summation can be calculated by regarding that the simultaneous occurrence of the number of bits equal to or more than “r” is stochastically negligible.

Also, it is required that a probability of SF/SD detection therefrom is calculated to carefully obtain a probability of a false detection and false release at a rate differing by one digit.

From this calculation result, the number of collected frames that meets a detecting (releasing) condition is derived to prepare the table shown in FIG. 2. FIG. 3 shows the table of FIG. 2 converted into graph form on a double logarithmic graph and shows a linearity as a whole.

As described above, the SF/SD detecting portion 6 having received the estimated maximum error cycle BER_max obtained by the conversion table 5 compares the estimated maximum error cycle BER_max with the preset alarm detecting threshold Th_d and the alarm releasing threshold Th_r to determine the relationship therebetween. A comparison example in this case is shown in the following Table 1.

TABLE 1

Namely, as shown in Table 1, when the output BER_max from the conversion table 5 gradually increases from a value lower than the threshold Th_r, namely when the error state of the input data worsens, an alarm signal is not generated from the SF/SD detecting portion 6 in states (2) and (3) where BER_max<alarm detecting threshold Th_d, but the SF/SD detecting portion 6 outputs the alarm ALM in a state (1) where BER_max≦alarm detecting threshold Th_d.

Thereafter, when the error state of the input data is improved and the output BER_max of the conversion table 5 gradually decreases, the alarm detected state continues in the states (1) and (2) where the BER_max is equal to or less than the alarm releasing threshold Th_r, but the SF/SD detecting portion 6 releases the alarm on the supposition that the BER_max enters the state (3) at the time when the BER_max becomes smaller than the alarm releasing threshold Th_r.

Thus, without using the timer, the alarm signal is generated when the BER of the input data worsens, and the alarm is released when the BER returns to a favorable state.

On the other hand, the alarm release of the SF/SD detecting portion 6 is also performed by the error-free detecting portion 7. This is because in the alarm releasing operation, the error cycle measurement is disabled upon entering an error-free state, that is a state where an error does not occur in the input data. Therefore, the alarm is released when it becomes error-free in a predetermined time period.

Accordingly, the error-free detecting portion 7 is provided with an alarm ALM from the SF/SD detecting portion 6, so that the error-free detecting portion 7 is activated by the alarm signal ALM.

Also, while the cycle monitoring reference clock CLK is constantly provided, when the error cycle T_e outputted from the error cycle detecting portion 2 stays flat for a time period of an error cycle threshold T_eThr corresponding to the above-mentioned alarm releasing threshold Th_r (namely when a parity error is not included in the input data and the error cycle T_e is held due to an error pulse P_E from the parity check portion 1 being not generated), the error-free detecting portion 7 releases the alarm state of the SF/SD detecting portion 6 by generating an error-free signal E_f by regarding that it is an error-free state where the alarm state should be released. Also, the error cycle memory 3 is initialized by the error-free signal E_f. It is to be noted that the error cycle threshold T_eThr need not necessarily correspond to the alarm releasing threshold Th_r.

Embodiment [2]: FIG. 4

While the maximum error cycle T_emax is obtained by using the error cycle maximum value retrieving portion 4 in the above-mentioned embodiment [1], this embodiment [2] is different in that an error cycle average calculating portion 4a calculating an average T_eave substituted for the maximum value T_emax of the error cycles is used. Therefore, the output of the conversion table 5 assumes an estimated error rate BER_ave corresponding thereto.

Namely, while the above-mentioned embodiment [1] is premised on scattered errors, this embodiment [2] adopts an arithmetic average, thereby enabling the estimation of the BER by using smoothed error cycles where influence of distribution is reduced even when a burst error occurs.

Embodiment [3]: FIG. 5

This embodiment is different from the above-mentioned embodiment [1] in that an error cycle median calculating portion 4b for calculating a median T_ecen of the error cycles is substituted for the error cycle maximum value retrieving portion 4. Therefore, the output of the conversion table 5 assumes an estimated error rate BER_cen corresponding thereto.

Namely, while the above-mentioned embodiment [1] is premised on the scattered error, this embodiment enables an estimation of an error rate reducing the influence even in the case where a burst error occurs by calculating the median of the error cycle.

Embodiment [4]: FIG. 6

This embodiment is different from the above-mentioned embodiment [1] in that an error cycle deviation calculating portion 8 is added.

Namely, the error cycle deviation calculating portion 8 constantly obtains the deviation of the error cycles at the time of error detection with the read request R_req from the error cycle maximum value retrieving portion 4 for the error cycle data stored in the error cycle memory 3. When the alarm is generated from the SF/SD detecting portion 6, the error cycle deviation calculating portion 8 determines whether or not the deviation value at this time exceeds a burst detecting threshold Th_b and generates a burst flag when deviation value exceeds the burst detecting threshold Th_b.

Thus, it is made possible to determine whether or not the current alarm detected state is due to a burst state.

It is to be noted that this is an operation performed only when a burst error is detected, so that this operation is not activated upon alarm release.

The present invention is not limited by the above-mentioned embodiments, and it is obvious that various modifications may be made by one skilled in the art based on the recitation of the claims.

Claims

1. A BER monitoring circuit comprising: a first means detecting error cycles of input data;a second means detecting a maximum value from among the error cycles;a third means converting the maximum value into a corresponding estimated error rate; anda fourth means generating an alarm when the estimated error rate exceeds an alarm detecting threshold.

2. The BER monitoring circuit as claimed in claim 1, wherein the fourth means includes means releasing the alarm when the estimated error rate assumes equal to or less than an alarm releasing threshold.

3. The BER monitoring circuit as claimed in claim 2, further comprising a fifth means being activated when the alarm is generated and releasing the alarm when a time period for which the error cycles stay flat exceeds a cycle corresponding to the alarm releasing threshold.

4. The BER monitoring circuit as claimed in claim 1, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

5. The BER monitoring circuit as claimed in claim 2, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

6. The BER monitoring circuit as claimed in claim 3, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

7. A BER monitoring circuit comprising: a first means detecting error cycles of input data;a second means detecting an average of the error cycles;a third means converting the average into a corresponding estimated error rate; anda fourth means generating an alarm when the estimated error rate exceeds an alarm detecting threshold.

8. The BER monitoring circuit as claimed in claim 7, wherein the fourth means includes means releasing the alarm when the estimated error rate assumes equal to or less than an alarm releasing threshold.

9. The BER monitoring circuit as claimed in claim 8, further comprising a fifth means being activated when the alarm is generated and releasing the alarm when a time period for which the error cycles stay flat exceeds a cycle corresponding to the alarm releasing threshold.

10. The BER monitoring circuit as claimed in claim 7, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

11. The BER monitoring circuit as claimed in claim 8, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

12. The BER monitoring circuit as claimed in claim 9, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

13. A BER monitoring circuit comprising: a first means detecting error cycles of input data;a second means detecting a median of the error cycles;a third means converting the median into a corresponding estimated error rate; anda fourth means generating an alarm when the estimated error rate exceeds an alarm detecting threshold.

14. The BER monitoring circuit as claimed in claim 13, wherein the fourth means includes means releasing the alarm when the estimated error rate assumes equal to or less than an alarm releasing threshold.

15. The BER monitoring circuit as claimed in claim 14, further comprising a fifth means being activated when the alarm is generated and releasing the alarm when a time period for which the error cycles stay flat exceeds a cycle corresponding to the alarm releasing threshold.

16. The BER monitoring circuit as claimed in claim 13, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

17. The BER monitoring circuit as claimed in claim 14, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

18. The BER monitoring circuit as claimed in claim 15, further comprising a sixth means constantly calculating a deviation of the error cycles and generating a burst flag when the deviation upon generation of the alarm exceeds a burst detecting threshold.

19. A BER monitoring circuit comprising: a first detector detecting error cycles of input data;a second detector detecting at least one of a maximum value, an average value, and a median value from among the error cycles;a converter converting the value into a corresponding estimated error rate; anda generator generating an alarm when the estimated error rate exceeds an alarm detecting threshold.

20. A BER monitoring method comprising: a first step of detecting error cycles of input data;a second step of detecting at least one of a maximum value, an average value, and a median value from among the error cycles;a third step of converting the value into a corresponding estimated error rate; anda fourth step of generating an alarm when the estimated error rate exceeds an alarm detecting threshold.