Multiplex communications – Communication techniques for information carried in plural... – Combining or distributing information via time channels
Reexamination Certificate
1998-04-09
2001-07-17
Olms, Douglas (Department: 2661)
Multiplex communications
Communication techniques for information carried in plural...
Combining or distributing information via time channels
C375S355000, C375S371000
Reexamination Certificate
active
06262999
ABSTRACT:
FIELD OF THE INVENTION
This invention relates generally to the field of asynchronous data networks, for example, ATM (asynchronous transfer mode networks), and in particular a method of clock recovery over asynchronous networks.
BACKGROUND OF THE INVENTION
ATM networks are essentially asynchronous in nature. In some applications, some means of recovering timing information as the data is transported through the network must be provided. For example, when a TDM link, such as a primary rate link, is connected to the output of an ATM switch, the timing information must be recovered to generate the primary rate clock of the source link.
One known method of recovering timing information across ATM networks is the Synchronous Residual Time Stamp (SRTS). This process, which assumes the network is timed to a master clock, is defined in ITU Recommendation I.363. The ATM network clock is taken and divided by 2[k], where k=0.1 . . . 11. k is chosen to produce a frequency near the service bit-rate of the source link (greater than but less than twice the rate). The difference between the service bit-rate and the network related frequency is produced and the 4 LSBs of this difference are transmitted using 4 CSI bits from 8 successive ATM cells. These 4 bits comprise the SRTS. Both transmitter and receiver carry out this process and by comparing the locally produced SRTS with the received far-end SRTS, the local clock can be synchronized to the remote clock. Section
6
of the Bellcore document, GR-1113-CORE describes in detail a method of implementing STRS clock recovery.
In a known method of primary rate clock recovery (See Integrated Telecom Technology, Inc. AAL1 SAR Processor WAC-
021_B, User's Manual, Data Sheet,
1996) the 1.544 MHz nominal primary rate clock is recovered using the Synchronous Residual Time Stamp (SRTS) method. The SRTS value transported over the ATM fabric is extracted by the SAR device at the receiver. This device compares the received SRTS value with a locally generated SRTS value derived from the receiver clock and outputs a signed 4-bit correction value having a value between −8 and +7.
The correction value is used to control the number of pulses of a 1.55520 MHz clock to be multiplexed with a fixed number of 1.495384 MHz clock pulses. The suggested ratio is 158+n 1.5552 MHz pulses to 35 1.495384 MHz pulses in a period of 193+n pulses, where n=the SRTS correction value. Thus, if the correction value is zero (signifying a nominal 1.544 MHz clock), the output clock will have a frequency deviation (fd) from nominal of +11,200/−48,616 Hz. The frequency of modulation from nominal fm) depends on the implementation with a non-sinusoidal 8 kHz fm at the low end, which will vary slightly depending on the magnitude and sign of the correction value. This represents intrinsic jitter, which can be defined in peak-peak Unit Intervals (UI) as Jp-p=fd/&pgr;*fm, where fm is a sinusoidal modulating frequency and fd is the frequency of deviation of the carrier frequency.
The use of a correction value implies that the output clock is averaged over time to arrive at the final locked frequency. This function creates low frequency jitter. Any averaging jitter below 6 Hz (more properly referred to as wander) will pass through the jitter attenuator unattenuated. The SRTS method itself also has the characteristic of waiting time jitter, which, by nature, is 1 UI. The above prior art arrangement is relatively complicated to implement.
The prior art method of clock recovery requires complex logic circuitry to implement.
An object of the invention is to provide a clock recovery method that can be implemented with simple logic circuitry.
SUMMARY OF THE INVENTION
According to the present invention there is provided a method of generating a corrected clock signal from a signed correction value derived from cell s carrying timing information in an asynchronous network, comprising the steps of deriving a decrement/increment indicator from the sign of the received correction value, deriving the true magnitude from the received correction value, and incrementing or decrementing an output signal in a predetermined period accordance with said decrement/increment indicator and said true magnitude.
In a preferred embodiment, the increment/decrement unit indicator enables a divide-by-16 counter in such a manner as to increase the period of an LSB (Lowest Significant Bit) by 25% when the final frequency is to decrease, or to decrease the period of an LSB by 25% for an increase in output frequency. The MSB (Most Significant Bit) of the divide-by-16 counter provides a nominal DS
1
clock output for a primary rate T
1
link. This output can be passed through an external jitter attenuator
8
before being presented as a clock.
REFERENCES:
patent: 4961188 (1990-10-01), Lau
patent: 5260978 (1993-11-01), Fleischer et al.
patent: 5608731 (1997-03-01), Upp et al.
patent: 5742649 (1998-04-01), Muntz et al.
patent: 5812618 (1998-09-01), Muntz et al.
patent: 5822383 (1998-10-01), Muntz et al.
patent: 5838749 (1998-11-01), Casper et al.
(Marks & Clerk)
Alcatel Canada Inc.
Olms Douglas
Phunkulh Bob A.
LandOfFree
Generation of primary rate clocks from correction values... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Generation of primary rate clocks from correction values..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Generation of primary rate clocks from correction values... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2500839