Title of Invention

"A SYSTEM AND METHOD FOR GENERATING A SYSTEM CLOCK SIGNAL"

Abstract

A system for generating a central clock signal which may periodically be locked to a selected one of a plurality of external clock signals, each of said external clock signals being generated in phase-locked relation-ship with an external reference signal by a digital phase-locked loop comprising a numerically controlled oscillator (6), which generates the clock signal of the loop by division of a common system clock signal, fed to all loops, by a factor which is determined by a digital error signal indicating a phase difference between the clock signal of the loop and the external reference signal of the loop, characterized in that the system comprises a central numerically controlled oscillator (7) and is adapted to transfer the digital error signal from the selected phase-locked loop to the central numerically controlled oscillator (7), and that the numerically controlled oscillator (7) is adapted to generate the central clock signal by division of the common system clock signal by a factor which is determined by the transferred digital error signal.

Full Text

The invention relates to a method of generating a central clock signal which may periodically be locked to a se¬lected one of a plurality of external clock signals. Each of the external clock signals is generated in phase-locked relationship with an external reference signal by a digital phase-locked loop comprising a numerically con-trolled oscillator/ which generates the clock signal of the loop by division of a common system clock signal, fed to all loops, by a factor which is determined by a digi¬tal error signal indicating a phase difference between the clock signal of the loop and the external reference signal of the The loop charactired in that the ciruit comprenis a untial numaricallly controlled oscillator and that the nummerically controlled oscillator is adopted to granted the central clock invention also relates to a corresponding system by division of the common system clock signal synchronous telecommunications systems, such as e.g an SDH network (Synchronous Digital Hierarchy), consist of a large number of network elements which each comprise a system clock generator which generates an internal clock signal to control the network element. An SDH network element may e.g. have an internal system clock generator of 77.76 MHz. This will typically be phase-locked to an 8 kHz signal derived from the received data, said 8 kHz corresponding to the frame frequency of the transmitted communications signals. For example, it is possible to lock to data received from another network element, or to one of the lower order signals, e.g. a PDH signal, which the network element receives with a view to introducing these into the SDH structure. Thus, each network element has its own system clock generator, typically a phase-locked loop (PLL). by a tutor a which is determind by the transferred digital error signal The individule network elements,which may e.g be multi- plexers or Add/Drop multiplexers, will frequently serve to multiplex a large number of plesiochronous channels, e.g. 2 Mbits/s channels, into the SDH system. In prin¬ciple, each of these 2 Mbits/s channels will have its own clock which is independent of the other channels and of the SDH system. Although the clock frequency of each channel will be close to the others, since all of them are produced as multiples of 8 kHz, but not necessarily the same 8 kHz source, their mutual phase may vary freely. In such a network element it is desirable to have a cen¬tral clock signal which is phase-locked to an arbitrary one of the plesiochronous channels to which it is con¬nected. The invention relates to such locking, but may also be used in other applications in which there is a need for a central clock signal which is phase-locked to an arbitrary one of a plurality of clock signals. Typically, each input gate or module, which receives a plesiochroncus channel, will be provided with a phase-locked loop which generates a clock signal corresponding thereto on the basis of the received data. It is known to transfer each of these clock signals to a central clock unit which can then select an arbitrary one of these clock signals as the central clock signal. However, it takes a large bandwidth to transfer the many clock sig-nals from the individual modules to the central clock unit, and it moreover means that a phase jump will typi-cally occur in the central clock signal when the central clock unit switches from one clock signal to another. Such a phase jump may cause the network to get out cf synchronization, and is therefore undesirable. Accordingly, an object of the invention is to proved a method of the type stated in the opening paragraph, which requires a smaller bandwidth between the individual mod¬ules and the central clock unit, and which enables switching from one module to another without a phase jump in the central clock signal. This is achieved according to the invention in that the digital error signal from the selected phase-locked loop is transferred to a central numerically controlled oscil¬lator, which generates the central clock signal by divid¬ing the common system clock signal by a factor determined by the transferred digital signal. When just the digital error signal is transferred, a con-siderable saving in bandwidth is achieved, as this varies with a much smaller frequency than the clock signal it-self. Further, the central numerically controlled oscil-lator will serve as a slave of the corresponding oscilla- ter in the selected loop, and when switching to another loop it will merely instead receive the error signal from the new loop, so that a phase jump will not occur in its Output signal which constitutes the central clock signal. A further saving in bandwidth may be achieved when, as stared in claim 2, the digital error signal froiu the se-lected loop is integrated over a period of time before it is transferred to the central numerically controlled os-cillator. As mentioned, the invention moreover relates to a circuit for generating a central clock signal of the stated type. When the circuit comprises a central numerically con¬trolled oscillator and is adapted to transfer the digital error signal from the selected phase-locked loop to the central numerically controlled oscillator, while the nu- merically controlled oscillator is adapted to generate the central clock signal by division of the common system clock signal by a factor which is determined by the transferred digital error signal, then, as mentioned above, a saving in bandwidth is achieved and no phase jump occurs when switching to another loop. As stated in claim 4, a further saving in bandwidth may be achieved in that each of the phase-locked loops moreo-ver comprises means for integrating the digital error signal of the loop over a period of time before it is transferred to the central numerically controlled oscil-lator. As stated in claim 5, these means may expediently comprise a counter capable of counting the number of times the numerically controlled oscillator has divided the system clock signal by a specific factor within a given period of time. Then, it suffices to transfer the count once for each period. the circuit will be particularly useful for an SDH system. when, as stated in claim 6, the common system clock sig-nal has a frequency of approximately 77.76 MHz. When, as stated in claim 7, the clock signal of each of the phase-locked loops has a frequency of approximately 2.048 MHz., the system is particularly useful for 2.048 Mbits/s chan-nelr,, v;hlch are mainly used in Europe, while, when the clock signal of the phase-locked loops has a frequency of approximately 1.544 MHz, as stated in claim 8, it is par¬ticularly useful for 1.544 Mbits/s channels, which are mainly used in North America. The invention will now be described more fully below with reierence to the drawing, in wh fig. 1 show an embodiment of the invention in which a digital error signal is transferred from a module to a central numerically controlled oscillator, and fig. 2 shows a variant of an external module. Fig. 1 shows an example of an embodiment of the inven¬tion. In the shown example, the invention is applied in a network element in an SDH network in which a plurality of 2 Mbits/s channels are to be introduced into the SDH sys¬tem. A system clock generator 1, which generates a clock signal of 77.76 MHz, is placed on a central module 2, while for each 2 Mbits/s channels there is an external module, of which the two modules 3 and 4 are shown here. Each of the external modules contains a clock regenera-tion circuit, which consists of a digital phase-locked loop having a phase detector 5 and a numerically con-trolled oscillator (NCO) 6. This circuit regenerates a clock signal of 2.048 HHz from an arriving data signal (e.g. EXTI) of 2 Mbits/s. The numerically controlled os-cillator is designed as a programmable divider which di-vides the system clock signal of the 77.76 MHz. Theoreti-cally, the division ratio should be 77.76/2.048 = 37.97, and the divider 6 is therefore adapted to be capable of dividing by a factor of 37, 38 or 39 depending on a con trol signal which may assume the values {-1,0,1}. The phase detector 5 is therefore adapted to emit one of these values depending on the phase difference between the arriving signal and the regenerated clock signal, and the control signal is therefore also called an error sig-nal . According to the prior art, generation of a central clock signal of 2.048 MHz phase-locked to an arbitrary one of the many channels would require feeding each of the re-generated 2.048 MHz clock signals to the central module 2, where one of them would be selected as the instan-taneous central clock signal. However, it requires a con-siderable bandwidth to transfer the many 2 MHz signals, and moreover a phase jump would occur in the central clock signal when switching from one channel to another. Instead, in the shown embodiment of the invention, a nu-merically controlled oscillator or divider 7, which may be quite similar to the oscillators 6 placed on the ex-ternal modules, is arranged on the central module 2, and then just the error signal from each external module is transferred to the central divider 7. The central divider 7 will thus follow the divider 6 on the selected module closely, as they receive exactly the same error signal. This means that the two dividers or oscillators have ex-actly the same frequency, but not necessarily the same phase. A selector 8 connects the selected error signal to the divider 7. When switching from one module to another, the divider 7 merely receives a new error signal, but will otherwise not be affected, and thus no phase jump will occur in the central clock signal. The above description relates to a situation in which a plurality of 2 Mbits/s channels are introduced into an SDH system. The same circuit may also be used when e.g. 1.5 Mbits/s channels are involved. In that case, the re-generated clock signal is of 1.544 MHz, and the division ratio of the programmable divider should therefore be 77.76/1.544 = 50.36 in theory. In this case, the divider is therefore arranged to be capable of dividing by a fac-tor of 50 or 51 depending on the error signal, which may here assume the values {0, 1}. fig. 2 show a variant of the external module 3. Instead of currently transferring the error signal to the central module 2 like in fig. 1, the values are here accumulated in an accumulator 9. The accumulator may e.g. be an 8-bit counter which counts the number of times the programmable counter 6 has counted to 37 or 39 (for 2 Mbits/s) or 50 (for 1.5 Mbits/s) within a specific period of time. In the example, the count, also called synchronization pointer (SP) , is transferred to the central module 2 once for each SDH frame, and the counter 9 is therefore con¬trolled by an 8 kHz signal, which transfers SP and then resets the counter. Additional bandwidth is saved hereby, since it is just necessary to transfer a byte (SP) for each frame period. In the 2 Mbits/s case, the count (SP) of the counter 9 is increased by 1 (i.e. SP = SP + 1) when the divider 6 di¬vides by 37. It is maintained unchanged (i.e. SP = SP + 0) when the divider 6 divides by 38, and it is reduced by one 1 (i.e. SP = SP - 1; when the divider 6 divides by 39. As the theoretical division ratio 37.97 is close to 38, the divider 6 will divide by 38 in the vast majority of cases, and the count will therefore assume a value as low as 8 for a period at nominal frequencies, and, when the frequencies vary, it will be within the range 0-16. lr. this case, a 4-bit counter will thus be sufficient. In case of a 1.5 Mbits/s channel, the count (SP) of the counter 9 is increased by 1 (i.e. SP = SP + 1) when the divider 6 divides by 50, and it is maintained unchanged (i.e. SP - SP + 0) when the divider 6 divides by 51. Here the count will assume the value 123 for a period at nomi-nal frequencies, and a counter of 7 or 8 bits has to be used. on thecentral module 2, the SP value is fed from the se lected module to the divider 7 which, in the subsequent period, adjusts its division ratio to the SP value just received. Although a preferred embodiment of the present invention has been described and shown, the invention is not re¬stricted to it, but may also be embodied in other ways within the scope of the subject-matter defined in the ap¬pended claims. CLAIM: 1. A system for generating a central clock signal which may periodically be locked to a selected one of a plurality of external clock signals, each of said external clock signals being generated in phase-locked relation-ship with an external reference signal by a digital phase-locked loop comprising a numerically controlled oscillator (6), which generates the clock signal of the loop by division of a common system clock signal, fed to all loops, by a factor which is determined by a digital error signal indicating a phase difference between the clock signal of the loop and the external reference signal of the loop, characterized in that the system comprises a central numerically controlled oscillator (7) and is adapted to transfer the digital error signal from the selected phase-locked loop to the central numerically controlled oscillator (7), and that the numerically controlled oscillator (7) is adapted to generate the central clock signal by division of the common system clock signal by a factor which is determined by the transferred digital error signal. 2. A system as claimed in claim 1, wherein each of the phase-locked loops moreover comprises means (9) for integrating the digital error signal of the loop over a period of time before it is transferred to the central numerically controlled oscillator (7). 3. A system as claimed in claim 2, wherein said integrating means (9) comprise a counter capable of counting the number of times the numerically controlled oscillator (6) has divided the system clock signal by a specific factor within a given period of time. 4. A system as claimed in claims 1-3, wherein the common system clock signal has a frequency of approximately 77.76 MHz. 5. A system as claimed in claims 1-4, wherein the clock signal of each of the phase-locked loops has a frequency of approximately 2.048 MHz. 6. A system as claimed in claims 1-4, wherein the clock signal of each of the phase-locked loops has a frequency of approximately 1.544 MHz. 7. A method of generating a central clock signal which may periodically be locked to a selected one of a plurality of external clock signals, each of said external clock signals being generated in- phase-locked relationship with an external reference signal by a digital phase-locked loop comprising a numerically controlled oscillator (6), which generates the clock signal of the loop by division of a common system clock signal, fed to all loops, by a factor which is determined by a digital error signal indicating a phase difference between the clock signal of the loop and the external reference signal of the loop, charact e r i zed in that the digital error signal from the selected phase-locked loop is transferred to a central numerically controlled oscillator (7), which generates the central clock signal by division of the common system clock signal by a factor determined by the transferred digital error signal. 8. A method as claimed in claim 1, wherein the digital error signal from the selected loop is integrated over a period of time before it is transferred to the central numerically controlled oscillator (7). 9. A system for generating a central clock signal substantially as herein described with reference to the accompanying drawings.

Documents:

3467-del-1997-abstract.pdf

3467-del-1997-claims.pdf

3467-del-1997-correspondence-others.pdf

3467-del-1997-correspondence-po.pdf

3467-del-1997-description (complete).pdf

3467-del-1997-drawings.pdf

3467-del-1997-form-1.pdf

3467-del-1997-form-13.pdf

3467-del-1997-form-19.pdf

3467-del-1997-form-2.pdf

3467-del-1997-form-3.pdf

3467-del-1997-form-4.pdf

3467-del-1997-form-6.pdf

3467-del-1997-gpa.pdf

3467-del-1997-pct-210.pdf

3467-del-1997-pct-408.pdf

3467-del-1997-pct-409.pdf

3467-del-1997-petition-137.pdf

3467-del-1997-petition-138.pdf