WO1999059052A1

WO1999059052A1 - Method and apparatus for synchronisation of nodes

Info

Publication number: WO1999059052A1
Application number: PCT/SE1999/000613
Authority: WO
Inventors: Stellan Lundqvist; Lars Westberg
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 1998-05-11
Filing date: 1999-04-16
Publication date: 1999-11-18
Also published as: JP2002514877A; SE512034C2; AU4400899A; EP1078314A1; CA2331963A1; SE9801631L; SE9801631D0; CN1309786A

Abstract

The present invention relates to a method and apparatus for synchronising at least one node (N1-N4) in a net (n), which node (N1-N4) is used as an example for packet oriented transmission. A server (C), generating a stable frequency (fs), and the node (N1-N4) measures the number of cycles of the stable frequency (fs) respective the number of cycles of a local frequency (f1-f4), generated by the node (N1-N4). These measurements are started respective stopped at the instant of time when a phase of a signal, generated by a power network (PN) connected to the server (C) and the node (N1-N4), equals zero. The results of the measurements are compared and the node (N1-N4) is adjusted accordingly. Measurements can continuously be performed by the server (C) and the node (N1-N4).

Description

Method and Apparatus for Synchronisation or Nodes

TECHNICAL FIELD OF THE INVENTION

The present invention relates to a method and apparatus in a network, wherein a node in the network generates a frequency, which needs to be synchronised with a reference frequency in order to obtain an exact frequency in the node.

DESCRIPTION OF RELATED ART

According to prior art a communication network, such as a mobile network or cellular network, constitutes cells, wherein each cell is equipped with a basestation. Each basestation operates on a set of radio channels with a respective carrier frequency, which channels are different from the channels used in neighbouring cells to avoid interference .

A group of basestations are controlled by a basestation controller (BSC) and a group of basestation controllers in turn are controlled by a switching centre (MSC) . The switching centre connects different networks with each other like mobile networks (GSM, AMPS, NMT, UMTS) and the public switched telephone network (PSTN) .

Generally a transmission between the basestation controller and the basestations in the network uses a PCM-link (Pulse Coded Modulation link) .

The frequency of the PCM-link is used by the basestations as a reference to synchronise a local frequency in the basestations. The frequency of the PCM-link has a very high long term stability, with a deviation that is normally less than 0,05 ppm (parts per million) .

Each basestation is connected to a power network, which network supplies the basestation with power. The local frequency, generated by a local oscillator situated in the basestation, is normally used as a local reference to generate the carrier frequencies belonging to the different radio channels of the basestation.

In case the transmission between the network and a basestation constitutes of information sent in packets, also called packet oriented information, no PCM-link is used and therefore the frequency of the PCM-link can not be used by the basestation to synchronise the local frequency in the basestation. An example of packet oriented information is information transmitted over an Intranet/Internet. In this case, an IP-network (Internet Protocol-network) is used instead of the PCM-link to transport the packet oriented information between corresponding units.

A known method to synchronise the local frequency in a basestation when transmitting packet oriented information is to use high stable oven oscillators in each basestation, which oscillators synchronise the local oscillators in the corresponding basestation.

A problem with this method is that the oven oscillators are large and expensive and they need to be calibrated at regular intervals, which is an expensive operation.

The patent document US 4, 602, 340 describes a system for distributing values of coded time or other information signals through the existing electrical gridwork of a facility by reading the values of a clock in a place and formatting the read information to a suitable serial form. The formatted information is sent over the electrical gridwork by a form of modulation via the grounded leg in the gridwork.

SUMMARY OF THE INVENTION

The problem dealt with by the present invention is to synchronise at least one frequency in at least one node in a net, also called network. One intention of the invention is thus to synchronise at least one frequency in at least one node in the network.

The problem is solved essentially by using the phase of a signal generated by a power network, connected to the node, as a reference to define a stable time interval during which time interval at least one measurement is performed of the frequency in the node to be synchronised.

Further at least one server generating a stable frequency is used, which server is connected to the power network, and the server performs a measurement of the stable frequency during the same time interval as the measurement of the frequency is performed in the node to be synchronised.

More specifically, the problem is solved in the following manner .

The server and the node performs under the well defined time interval at least one measurement of the number of cycles of the stable frequency generated by the server respective the number of cycles of a local frequency generated by a local oscillator in the node.

When a measurement is about to start, the server sends a message to the node comprising information about the length of the measurement, i.e. the time interval during which the measurement shall be performed. A measurement in both the server and in the node is started respective stopped when the phase of the signal, generated by the power network PN, equals a predefined threshold value.

The resulting calculated number of cycles of the stable frequency in the server and the local frequency in the node are compared and the oscillator in the node is adjusted accordingly.

In an alternative embodiment of the invention measurements are continuously performed in the server and in the node. One advantage afforded by the invention is that it proposes a simple and general method to synchronise a frequency in a node.

Another advantage is that the method according to the invention works even if there is a phase difference between the server and the node to be synchronised.

The invention will now be described in more detail with reference to exemplifying embodiments thereof and also with reference to the accompany drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a schematic view of a system comprising different nets, according to the invention,

Figure 2a and 2b are flowsheets illustrating a method according to the invention, and

Figure 3 is a block schematic illustrating a node and a clock server according to the invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS

Figure 1 shows a schematic view of a number of nodes N_x-N₄ in a net n, also called network. The network n is for example an internet network or a cellular network like GSM, NMT, AMPS, UMTS. In Figure 1 is as an example shown four nodes N].-N₄, but of course more or less nodes can be located in the network n.

Each node Nχ-N₄ is connected to a packet network IP, which in turn is connected to a switching centre MSC. The packet network IP is for example an internet protocol network.

The switching centre MSC connects the network n with other networks n₂, n₃ and the public switched telephone network PSTN, as shown in Figure 1.

Each node Nχ-N₄ is also connected to a power network PN, which power network PN supplies the nodes N_!- ₄ with power. A local oscillator L0ι-L0 is situated in each node Nχ-N₄, which local oscillator LOι-L0₄ generates a local frequency of for example 10 MHz. The local frequency in turn is used as a reference to generate a number of carrier frequencies in the corresponding node N_x-N₄. A carrier frequency corresponds for example to a radio channel in the node N_α-N .

Packet transmission units PC are as an example connected to the nodes Nχ-N₄ via the packet network IP as shown in Figure 1. The packet transmission units PC are used for sending packet oriented information to a corresponding node N_x-N₄ in the network n.

The packet transmission units PC in Figure 1 illustrates as an example a way to send packet oriented information to a corresponding node Ni-N₄. Packet oriented information can also be sent between the nodes N_x-N₄ and further between the switching centre MSC and a corresponding node N]_-N₄, but this is not described in this example. The invention is applicable to all kind of packet oriented transmissions to and from the nodes Nχ-N₄ in the network n.

Further, packet oriented transmission is described in the examples below to illustrate the invention, but the method according to the invention is also applicable to other kinds of transmissions to and from a corresponding node Nχ-N₄ to be synchronised. For example the invention could be applied for ATM transmission (Asynchronous Transfer Mode) or for SDH/PDH transmission (Synchronous Digital Hierarchy/Plesiochronous Digital Hierarchy) in cases where the clock stability is poor.

A clock server C, with a stable reference frequency f_s, is connected to the power network PN and the packet network IP. The clock server C is for example a personal computer equipped with a GPS receiver (Global Positioning System) .

A method according to the invention is described below in association with Figure 1, 2a, 2b and the above examples. The power network PN generates a signal with a frequency of for example 50 or 60 Hz, and the instant of time when the phase of this signal equals zero is in this example used as a reference to start and stop both a measurement of the reference frequency f_s in the clock server C and of a local frequency fι~f₄ in a node Nχ-N₄ in the network n. The mentioned phase is hereby used as a reference to define a stable time interval that start and end when the phase of said signal equals zero, during which time interval the measurement is performed of the reference frequency f_s in the clock server C and of a local frequency f_λ- f₄ in a node Ni-N .

This is done in order to synchronise the local frequency fι~f in the corresponding node N_!-N₄.

The phase of the mentioned signal does not necessarily have to equal zero in order to define the start and the end of the stable time interval. A threshold value close to zero or a threshold value comprising a hysteresis function could also be used according to the invention as a reference to trigger a start and a stop of the measurement. For simplicity the threshold value is equal to zero in the examples described below.

Figure 2a illustrates a flowsheet of a method according to the invention.

It is assumed in the following example that the node N_x-N₄, the power network PN, the packet network IP, the clock server C, the packet transmission units PC and the switching centre MSC are located as described above.

One or several of the packet transmission units PC are assumed to be communicating with a first node N_x, wherein a local frequency f_x in the first node N_x is synchronised as will be described in this example. The first node N_x is located in the network n as shown in Figure 1 and connected to the packet network IP and the power network PN. Of course the method according to the invention could be applied to more than one node at a time, i.e. more than one node Nj_-N₄ can be synchronised at the same time by letting the clock server C send start messages and end reports as described below to more than one node at a time. For clarity the first node Ni is synchronised as an example below.

In step 101, the clock server C sends a start message to the first node N_x via the packet network IP. The start message comprises information of the time interval, i.e. for how many cycles of the signal generated by the power network PN, a measurement of the local frequency f_x in the first node N_l as well as of the reference frequency f_s in the clock server C, shall be performed.

For purposes of illustration, the mentioned measurement time interval may be 180000 cycles of the signal generated by the power network PN.

When the clock server C in step 103 detects that the phase of the signal, generated by the power network PN, equals zero, a first measure of the reference frequency f_s in the clock server C is performed in step 105.

The clock server C performs a second measure of the reference frequency f_s in step 109 at the end of the mentioned time interval when the measurement of the reference frequency f_s shall end according to step 107. The second measure is performed when the phase of the signal, generated by the power network PN, equals zero and at the end of the mentioned time interval.

In step 111, the clock server C calculates the difference between the second and the first measure of the reference frequency f_s, wherein the number of cycles of the reference frequency f_s that has passed during the mentioned time interval is obtained. This result is included in an end report, which end report is sent to the first node Ni via the packet network IP in step 113.

An example of performing the measuring of the number of cycles of the reference frequency f_s, generated by a reference oscillator situated in the clock server C, that has passed during the mentioned time interval is described below. In this example it is assumed that a counter (not shown in the Figure) is connected to the reference oscillator. The counter counts the number of cycles of the reference frequency f_s.

When the first measure of the reference frequency f_s in the clock server C in step 105 is performed as described above, it is actually the value of the counter, connected to the reference oscillator, that is checked. This value, called first value below, is registered in for example a register.

Similarly the value of the counter is checked when the second measure of the reference frequency f_s in the clock server C in step 109 is performed as described above. This value, called second value below, is also registered in the register.

The first and the second value of the counter is checked when the phase of the signal, generated by the power network PN, equals zero, as described above i.e. at the start and at the end of the mentioned time interval.

The clock server C in step 111 calculates the difference between the second and the first value of the counter, wherein the number of cycles of the reference frequency f_s that has passed during the mentioned time interval is obtained.

In step 115, the first node N_x receives the start message, marked with Ai in figure 2a, which was sent from the clock server C in step 101. The first node N_x detects in step 117 when the phase of the signal, generated by the power network PN, equals zero. In step 119, the first node N_x then performs a first measure of the local frequency f_{l r} generated by the local oscillator LOi in the first node N_l7 when the mentioned phase of the signal is detected to equal zero.

The first measure of the local frequency f_x is as an example performed by checking the value of a counter (not shown in the Figure) , connected to the local oscillator LOi in the first node N_x. This value, called first value below, is registered in a register. The counter counts the number of cycles of the local frequency fi.

The first node _x performs a second measure of the local frequency fi in step 123 at the end of the mentioned time interval when the measurement of the local frequency fi shall end, according to step 121. The second measure is performed when the phase of the signal, generated by the power network PN, equals zero and at the end of the mentioned time interval.

The second measure of the local frequency fi is as an example performed by checking the value of the mentioned counter, connected to the local oscillator LOi. This value, called second value below, is registered in the mentioned register.

In the next step 125, the first node N_x calculates the difference between the second and the first measure of the local frequency fj_.. More specifically it is the difference between the second and the first value of the counter that is calculated, wherein the number of cycles of the local frequency f_x that has passed during the mentioned time interval is obtained. The first node N_x receives in step 127 the end report from the clock server C.

The first node N_x in step 129 compares the result in the end report from the clock server C with the result of the number of cycles of the local frequency f_x that has passed during the mentioned time interval.

If the result from the clock server C is different from the result in the first node _x according to step 131, then the first node _x in step 133 adjusts the local oscillator LOi in the first node N_x. Adjusting the local oscillator LOi is done by changing the voltage level that controls the local oscillator LOi so that the local frequency f_x equals the reference frequency f_s in the server C.

In another example according to the invention several measurements of the local frequency f_x in the first node N_x are performed to continuously synchronise the local frequency f_x in the first node N_x. This is described below in association with Figure 2b.

In this example it is assumed that a new measurement of the local frequency f_x in the first node N_x, as well as of the reference frequency f_s in the clock server C, is started every minute. In this example each measurement in the clock server C and in the first node N_x is performed for a time interval of 180000 cycles of the signal in the power network PN. Of course this is just an example and other lengths of the time interval are possible as well according to the invention.

Initially a first counter t is set to one in step 200. The first counter t increases by one every minute.

The method begins in step 201 with the clock server C sending a start message m_t to the first node N_x via the packet network IP. The first counter t is equal to one the first time a start message is sent from the clock server C, wherein the start message m_t is equal to m_x.

The start message comprises information about for how long time interval, i.e. for how many cycles of the signal generated by the power network PN, a measurement of the local frequency f_x in the first node N_X as well as of the reference frequency f_s in the clock server C, shall be performed. In this example this time interval is 180000 cycles of the power network PN generated signal, as mentioned above.

Further, the start message comprises information about the status of the first counter t so that the node N_x and the clock server C can keep track of the number of the corresponding measurement.

It is also possible to use predefined lengths of the mentioned time interval, which lengths are known by the server C and the first node N_X wherein only one start message m_t has to be used according to the invention in order to start a sequence of measurements in the clock server C and in the first node N_x. Start messages m_t to start every measurement are though described in these examples to better illustrate the method according to the invention.

When the clock server C in step 203 detects that the phase of the signal, generated by the power network PN, equals zero, the clock server C performs a first measure of the reference frequency f_s in the clock server C in step 205 for the measurement number t.

The first measure of the reference frequency f_s is as an example performed by checking the value of a counter (not shown in the Figure) , connected to the reference oscillator in the clock server C. This value, called first value below, is registered in a register. The counter counts the number of cycles of the reference frequency f_s.

In step 207 the first counter t is checked and if it has increased by one, i.e. one minute has passed, the clock server C continues to send a new start message m_t to the first node N_x via the packet network IP in step 201 as described above. For purposes of illustration, the new start message m_t is equal to m₂ the second time a start message is sent from the clock server C.

The clock server C keeps sending a new start message m_t to the first node N_x via the packet network IP according to the steps above every time t increases by one according to step 207.

The clock server C performs a second measure of the reference frequency f_s in step 211 for a measurement when the mentioned time interval for the corresponding measurement has ended and this measurement of the reference frequency f_s shall end according to step 209. The mentioned time interval is as described above 180000 cycles of the signal generated by the power network PN. The second measure is performed when the phase of the signal, generated by the power network PN, equals zero and 180000 cycles of this signal has passed.

The second measure of the reference frequency f_s is as an example performed by checking the value of the mentioned counter, connected to the reference oscillator. This value, called second value below, is registered in the mentioned register.

If the mentioned time interval has not ended according to step 209, the method returns to step 207 and the first counter t is checked. If the value of the first counter t has increased the method will return to step 201 where a new start message m_t will be sent from the clock server C.

If however the measurement time interval has ended, the clock server C calculates in step 213 the difference between the second and the first measure of the reference frequency f_s for a corresponding measurement. More specifically it is the difference between the second and the first value of the counter, connected to the reference oscillator, for the corresponding measurement that is calculated, wherein the number of cycles of the reference frequency f_s that has passed during the mentioned time interval is obtained for this measurement. This result is included in an end report, corresponding to the measurement, which end report is sent to the first node N_x via the packet network IP in step 215.

In step 217, the first node N_x receives the start message, marked with A₂ in figure 2b, for the measurement number t. The first node N_x detects in step 219 when the phase of the signal, generated by the power network PN, equals zero. In step 221, the first node N_x then performs a first measure of the local frequency f_x, generated by the local oscillator LO_x in the first node N_x, for the measurement number t when the mentioned phase of the signal is detected to equal zero.

The first measure of the local frequency f_x is as an example performed by checking the value of a counter (not shown in the Figure) , connected to the local oscillator LO_x in the first node N_x. This value, called first value below, is registered in a register. The counter counts the number of cycles of the local frequency f_x.

The first node N_x performs a second measure of the local frequency f_x for a measurement in step 225 when the mentioned time interval, corresponding to the measurement, has ended and the corresponding measurement of the local frequency f_x shall end, according to step 223. The second measure is performed when the phase of the signal, generated by the power network PN, equals zero and when 180000 cycles of this signal has passed.

The second measure of the local frequency f_x is as an example performed by checking the value of the mentioned counter, connected to the local oscillator LO . This value, called second value below, is registered in the mentioned register.

In the next step 227, the first node N_x calculates the difference between the second and the first measure of the local frequency f_x for a corresponding measurement. More specifically it is the difference between the second and the first value of the counter, connected to the local oscillator LO, for the corresponding measurement that is calculated, wherein the number of cycles of the local frequency f_x that has passed during the mentioned time interval, corresponding to the measurement, is obtained.

The first node N_x receives in step 229 the end report from the clock server C for the corresponding measurement. The first node N_x in step 231 compares the result in the end report from the clock server C with the result of the number of cycles of the local frequency f_x that has passed during the mentioned time interval for the corresponding measurement.

If the result from the clock server C is different from the result in the first node N_x for the corresponding measurement according to step 233, then the first node N_x in step 235 adjusts the local oscillator LO_x in the first node N_x, as described above.

The method continues to step 207 and the method is repeated as described above for the rest of the measurements performed in the clock server C and in the first node N_x.

A certain error of the phase difference in the power network PN may cause an error in the results of a measurement in the clock server C and in the first node N_x. The error of the phase difference depends on the distance between the clock server C and the first node N_x, and the topology of the power network PN. As an example an error of the phase difference of about 5 degrees, i.e. 1.5 %, in the power network PN is obtained between Enkδping and Ringhals in Sweden.

To minimise these measurement errors, the measurements in the clock server C and in the first node N_x are performed for a long time interval according to the above examples, for example under 180000 cycles of the signal in the power network PN. This corresponds to around sixty minutes (1 cycle=l/50 seconds => 180000 cycles=3600 seconds) .

An example of the accuracy of the method according to the invention is described below.

In this example it is assumed that the error of the phase difference in the power network PN is 2 degrees, i.e. 2/360=0,56%. This phase error give rise to a time difference of 0,0056x1/50=0,00011 seconds=0,ll ms between the clock server C and the first node N_x. This time difference must be counted twice since it will occur both at the start and at the stop of a measurement performed in the clock server C and in the first node N_x.

It is further assumed in this example that the time interval for a measurement is set to sixty minutes, i.e. 3600 seconds. The obtained accuracy is then 0,11x2 ms/3600=0, 0000000611*

The local oscillator L0_X in the first node N_x does not have to be adjusted after every comparison of results from a measurement as described in the above example. A mean value can be calculated of results from several measurements after these measurements have been performed in the clock server C and in the first node N_x.

The mean value is for example calculated in the first node N_x of the results from a number of measurements in the clock server C, which results are received by the first node N_x in the corresponding end reports, and of the results from the corresponding measurements in the first node N_x. These mean values are compared in the first node N_x, wherein the local oscillator L0_X in the first node N_x is adjusted accordingly.

The results from a number of measurements can be processed in some kind of filter, for example lowpass filter or Kalman filter, in the first node N_x to improve the method according to the invention also.

In another embodiment of the invention, the clock server C performs the treatment of the results obtained in the clock server C and in the node N_x-N to be synchronised, wherein the clock server C controls the corresponding node N_x-N₄. This is described below in association with the previous embodiments.

In this example it is assumed that the local frequency f_x in the first node N_x is to be synchronised. Of course several nodes N_x- N can be synchronised at the same time but for clarity the first node N_x is synchronised as an example below. It is also assumed in the following example that the node N_x- N₄ the power network PN, the packet network IP, the clock server C, the packet transmission units PC and the switching centre MSC are located as described above.

In the same way as described above, the clock server C and the first node N_x performs a measurement of the reference frequency f_s respective of the local frequency f_x during the stable time interval when a start message is sent to the first node N_x via the packet network IP. The time interval is defined to start and end when the phase of the signal, generated by the power network PN, equals zero as described above.

The result obtained of the number of cycles of the local frequency f_x that has passed during the mentioned time interval is included in a report, which report is sent from the first node N_x to the clock server C via the packet network IP.

The clock server C compares the result in the report from the first node N_x with the result of the number of cycles of the reference frequency f_s that has passed during the mentioned time interval.

If the result from the first node N_x is different from the result in the clock server C, then the clock server C sends an adjusting message to the first node N_x. The adjusting message comprise as an example of information about how much the local oscillator LO_x in the first node N_x should be adjusted in order to make the local frequency f_x be equal to the reference frequency f_s in the clock server C.

In the same way as described in association to Figure 2b, several measurements can be performed of the local frequency f_x in the first node N_x and of the reference frequency f_s in the clock server C to continuously synchronise the local frequency f_x in the first node N_x. Further, the clock server C can include a filtering function in which results from a number of measurements, performed in the clock server C and in the first node N_x, can be processed to improve the method according to the invention, wherein the local oscillator LO in the first node N_x is adjusted accordingly.

The node N comprise as an example a receiver 303, a transmitter 305, a measuring unit 307, a local oscillator LO 309, a memory unit 311, a filter unit 313, a calculating unit 315, a comparator unit 317, an adjusting unit 319, a power connector 321, a detector unit 322 and a control unit 323, which are the components shown in figure 3. The receiver 303 and the transmitter 305 are connected to an antenna 301. All the units are connected to each other by a databus 300 as shown in the figure.

The measuring unit 307 in the node N is used to measure the local frequency f_x, which local frequency f_x is used as a reference to generate several carrier frequencies in the node N.

The memory unit 311 stores the measures from the measuring unit

307, results from the calculating unit 315, and in some cases results from the calculating unit 415 in the server C described below.

The calculating unit 315 in the node N calculates for example a difference between two measures stored in the memory unit 311, which measures are obtained at the start and at the end of a measurement, and a mean value of results from several measurements stored in the memory unit 311.

The power connector 321 connects the node N to the power network PN.

The detector unit 322 in the node N detects when the phase of the signal, generated by the power network PN, equals zero. The detector unit 322 also performs calculation of the cycles of the local frequency f_x.

In some cases a comparator unit 317 and a filter unit 313 are situated in the node N, as shown in the figure.

The comparator unit 317 is used as an example for determining if the result of measurements sent from the clock server C, as described in the above examples, is different from the result of corresponding measurements stored in the memory unit 311 in the node N.

The adjusting unit 319 adjusts the local oscillator LO 309 when needed by changing the voltage that controls the local oscillator LO 309 so that the local frequency f_x equals the reference frequency f_s in the server C.

The units 303, 305, 307, LO 309, 311, 313, 315, 317, 319, 321, 322 and 323 are connected to the databus 300, through which the units communicate with each other. The control unit 323 in the node N controls the different units via the databus 300 and affects them to perform wanted operations according to the invention.

The clock server C comprise as an example a detector unit 405, a measuring unit 407, a reference oscillator 409, a memory unit 411, a calculating unit 415, a timer unit 416, a power connector 417 and a control unit 419, which are the components shown in figure 3. All the units are connected to each other by a databus 400 as shown in the figure.

The detector unit 405 detects when the phase of the signal, generated by the power network PN, equals zero. The detector unit 405 also performs calculation of the cycles of the reference frequency f_s, which stable reference frequency f_s is generated by the reference oscillator 409.

The measuring unit 407 in the clock server C is used to measure the reference frequency f_s. The memory unit 411 is used for example for storing measures from the measuring unit 407, results from the calculating unit 415, and in some cases results from the calculating unit 315 in the node N described above.

The calculating unit 415 calculates the difference between two measures stored in the memory unit 411, which measures are obtained at the start and at the end of a measurement, and a mean value of results from several measurements stored in the memory unit 411.

The timer unit 416 in the clock server C is used for handling the first counter t described above.

The power connector 417 connects the clock server C to the power network PN.

In some cases a comparator unit and a filter unit are situated in the clock server C but this is not shown in the figure.

In this case the comparator unit in the clock server C is used as an example for determining if the result of measurements sent from the node N, as described above, is different from the result of corresponding measurements stored in the memory unit 411 in the clock server C.

The units 405, 407, 409, 411, 415, 416, 417 and 419 are connected to the databus 400, through which the units communicate with each other. The control unit 419 in the clock server C controls the different units via the databus 400 and affects them to perform wanted operations according to the invention.

The signal, generated by the power network PN, comprise a voltage part and a current part. The instant of time when the phase of this signal equals zero is used above as a reference to start and stop a measurement in the clock server C and in the nodes N_x-N₄. More specifically it is the instant of time when the phase of the voltage part of this signal equals zero that is used above as a reference to start and stop a measurement in the clock server C and in the nodes N_x-N, i.e. as a reference to define the stable time interval during which a measurement is performed in the clock server C and in the nodes N_x-N₄.

The invention can be applied to all kind of nodes where an exact frequency is required. An example is a basestation.

The invention described above may be embodied in yet other specific forms without departing from the spirit or essential characteristics thereof. Thus, the present embodiments are to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing descriptions, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

Claims

1. Method for synchronising at least one node (N_x-N₄) in a net (n) , which node (N_x-N₄) is connected to a power network (PN) and a transmission network (IP) , wherein the method includes the following steps:

to perform a measurement of a stable frequency (f_s) generated by a server (C) , said server being connected to the power network (PN) and the transmission network (IP), during a predefined time interval, which time interval starts and ends when a phase of a signal, generated by the power network (PN), equals a threshold value;

to perform a measurement of a local frequency (f_x-f ) in the node (N_x-N₄) during the mentioned time interval;

to compare a result of the measurement obtained in the server (C) with a result of the measurement obtained in the node (N_x- N₄) ; and

to adjust the node (N_x-N₄) if the mentioned result in the server (C) is different from the mentioned result in the node (N_x-N₄) .

2. Method according to claim 1, wherein the method further includes the step to send a start message from the server (C) to the node (N_x-N₄) .

3. Method according to claim 2, wherein the method further includes the step to send an end report from the server (C) to the node (N_x-N₄) when the mentioned time interval has ended, which end report comprises the result of the measurement obtained in the server (C) .

4. Method according to claim 2, wherein the method further includes the step to send an end report from the node (N_x-N ) to the server (C) when the mentioned time interval has ended, which end report comprises the result of the measurement obtained in the node (N_x-N₄) .

5. Method according to claim 3, wherein the method includes the step to let the node (N_x-N₄) perform the comparison of the mentioned results obtained in the server (C) and in the node (N_x-N₄) .

6. Method according to claim 4, wherein the method includes the step to let the server (C) perform the comparison of the mentioned results obtained in the server (C) and in the node (N_x-N₄) .

7. Method according to claim 2, 3, 4, 5 or 6, wherein the method further includes the step to let the start message comprise information about the length of the mentioned time interval of the measurement in the server (C) and of the corresponding measurement in the node (N_x-N₄) .

8. Method according to claim 7, wherein the method further includes the step to let the time interval be a number of cycles of the signal, generated by the power network (PN) .

9. Method according to any of the claims 1-8, wherein the step to perform the measurement in the server (C) respective in the node (N_x-N₄) comprises the following steps:

to detect when the phase of the signal, generated by the power network (PN) , equals the mentioned threshold value;

to perform a first measure of the stable frequency (f_s) in the server (C) respective of the local frequency (f_x-f₄) in the node (N_x-N₄) when the phase of the signal, generated by the power network (PN) , is detected to equal the mentioned threshold value;

to detect when the phase of the signal, generated by the power network (PN) , equals the mentioned threshold value at the end of the mentioned time interval; and to perform a second measure of the stable frequency (f_s) in the server (C) respective of the local frequency (f_x-f₄) in the node (N_x-N ) at the end of the mentioned time interval when the phase of the signal, generated by the power network (PN) , is detected to equal the mentioned threshold value.

10. Method according to claim 9, wherein the method includes the step to let the result of the measurement in the server (C) be the number of cycles of the stable frequency (f_s) that has passed during the mentioned time interval, and to let the result of the measurement in the node (N_x-N₄) be the number of cycles of the local frequency (f_x-f₄) that has passed during the mentioned time interval.

11. Method according to any of the claims 1-10, wherein the method includes the step to let the local frequency (f_x-f₄) be generated by an oscillator (LO_x-L0₄) situated in the node (N_x- N₄) .

12. Method according to claim 11, wherein the step to adjust the node (N_x-N₄) comprise adjusting the oscillator (LO-L0 ) in the node (N_x-N) by changing a voltage that controls the oscillator (L0_X-L0₄) so that the local frequency (f_x-f₄) equals the stable frequency (f_s) in the server (C) .

13. Method according to any of the claims 1-12, wherein the method includes the step to let the mentioned threshold value be equal to zero.

14. Method for synchronising at least one node (N_x-N ) in a net (n) , which node (N_x-N ) is connected to a power network (PN) and a transmission network (IP) , wherein the method includes the following steps:

to perform a measurement of a stable frequency (f_s) generated by a server (C) , said server being connected to the power network (PN) and the transmission network (IP) , during a predefined time interval, which time interval starts and ends when a phase of a signal, generated by the power network (PN) , equals a threshold value;

to perform a measurement of a local frequency (f_x-f ) in the node (N-N₄) during the mentioned time interval;

to compare a result of a measurement obtained in the server (C) with a result of a corresponding measurement obtained in the node (N_x-N₄) ; and

to adjust the node (N_x-N₄) if the mentioned result in the server (C) is different from the mentioned result in the node (N_x-N₄) for the corresponding measurement.

15. Method according to claim 14, wherein the method further includes the step to send a start message (m_t) from the server (C) to the node (N_x-N₄) .

16. Method according to claim 15, wherein the start message (m_t) is sent from the server (C) to the node (N_x-N₄) at predefined intervals, wherein the start message (m_t) corresponds to a measurement.

17. Method according to claim 15 or 16, wherein the method further includes the step to send an end report from the server (C) to the node (N_x-N₄) when the mentioned time interval for a measurement has ended, wherein the end report comprise the result of the corresponding measurement obtained in the server (C) and an identity of the corresponding measurement.

18. Method according to claim 15 or 16, wherein the method further includes the step to send an end report from the node

(N_x-N₄) to the server (C) when the mentioned time interval for a measurement has ended, wherein the end report comprise the result of the corresponding measurement obtained in the node

(N_x-N₄) and an identity of the corresponding measurement.

19. Method according to claim 17, wherein the method includes the step to let the node (N_x-N₄) perform the comparison of the mentioned results obtained in the server (C) and in the node (N_x-N₄) .

20. Method according to claim 18, wherein the method includes the step to let the server (C) perform the comparison of the mentioned results obtained in the server (C) and in the node (N_x-N₄) .

21. Method according to claim 15, 16, 17, 18, 19 or 20, wherein the method further includes the step to let the start message (m_t) , corresponding to a measurement, comprise information about the length of the mentioned time interval of the measurement in the server (C) and of the corresponding measurement in the node (N_x-N₄) .

22. Method according to claim 21, wherein the method further includes the step to let the time interval be a number of cycles of the signal, generated by the power network (PN) .

23. Method according to any of the claims 14-22, wherein the step to perform the measurement in the server (C) respective in the node (N-N ) comprises the following steps:

to detect when the phase of the signal, generated by the power network (PN), equals the mentioned threshold value;

to perform a first measure of the stable frequency (f_s) in the server (C) respective of the local frequency (f_x-f₄) in the node (N_x-N₄) when the phase of the signal, generated by the power network (PN), is detected to equal the mentioned threshold value;

to detect when the phase of the signal, generated by the power network (PN), equals the mentioned threshold value at the end of the mentioned time interval; and

to perform a second measure of the stable frequency (f_s) in the server (C) respective of the local frequency (f_x-f₄) in the node (N_x-N₄) at the end of the mentioned time interval when the phase of the signal, generated by the power network (PN) , is detected to equal the mentioned threshold value.

24. Method according to claim 23, wherein the method includes the step to let the result of a measurement in the server (C) be the number of cycles of the stable frequency (f_s) that has passed during the mentioned time interval for the corresponding measurement, and to let the result of a measurement in the node

(N_x-N) be the number of cycles of the local frequency (f_x-f) that has passed during the mentioned time interval for the corresponding measurement.

25. Method according to claim 24, wherein the method further includes the following steps:

to process a result, from a predefined number of measurements performed. in the server (C) , in a filter;

to process a corresponding result, from the predefined number of corresponding measurements performed in the node (N_x-N₄) , in the filter;

to let the step of comparing the mentioned results obtained in the server (C) and in the node (N_x-N ) for a corresponding measurement include the step to compare the mentioned results from the filter for corresponding predefined number of measurements; and

to let the step of adjusting the node (N_x-N ) if the mentioned result in the server (C) is different from the mentioned result in the node (N_x-N₄) for the corresponding measurement comprise the step to adjust the node (N_x-N₄) if the mentioned results from the filter are different from each other for the corresponding predefined number of measurements.

26. Method according to claim 25, wherein the method includes the step to let node (N_x-N₄) include the filter.

27. Method according to claim 25, wherein the method includes the step to let the server (C) include the filter.

28. Method according to claim 26 or 27, wherein the method includes the step to let the filter be an averaging filter, a lowpass filter or a Kalman filter.

29. Method according to any of the claims 14-28, wherein the method includes the step to let the local frequency (f_x-f₄) be generated by an oscillator (LO_x-L0₄) situated in the node (N_x- N┬½) .

30. Method according to claim 29, wherein the step to adjust the node (N_x-N ) comprise adjusting the oscillator (LO_x-L0₄) in the node (N_x-N₄) by changing a voltage that controls the oscillator (LO_x-L0₄) so that the local frequency (f_x-f₄) equals the stable frequency (f_s) in the server (C) .

31. Method according to any of the claims 14-30, wherein the method includes the step to let the mentioned threshold value be equal to zero.

32. Apparatus in a net (n) , which apparatus is connected to a power network (PN) and a transmission network (IP) , wherein at least one measurement is performed of a local frequency (f_x-f) in the apparatus during a predefined time interval, which time interval starts and ends when a phase of a signal, generated by the power network (PN), equals a threshold value, wherein the apparatus comprises

receiver means (303) for receiving signals from the net

transmitter means (305) for transmitting signals to the net (n) ;

measuring means (307) for measuring the local frequency (f_x-f); oscillator means (LO 309) for generating the local frequency (f_x-f) ;

calculating means (315) for calculating a difference between two measures stored in a memory means (311) , which measures are obtained at the start and at the end of a measurement;

memory means (311) for storing measures performed by the measuring means (307) and results from the calculating means (315);

adjusting means (319) for adjusting the oscillator means (LO 309) ;

detector means (322) for detecting when the phase of the signal, generated by the power network (PN), equals the threshold value and for calculating the cycles of the local frequency (f_x-f₄); and control means (323) for controlling said means.

33. Apparatus according to claim 32, wherein the apparatus further comprises

filter means (313) for processing results from a number of measurements; and

comparator means (317) for determining if results of corresponding measurements are different.

34. Apparatus according to claim 32 or 33, wherein the calculating means (315) calculates a mean value of results from several measurements.

35. Apparatus according to claim 32, 33 or 34, wherein the mentioned time interval is a number of cycles of the signal, generated by the power network (PN) .

36. Apparatus according to claim 32, 33, 34 or 35, wherein said adjusting means (319) changes a voltage that controls the oscillator (LO_x-L0) .

37. Apparatus according to claim 36, wherein said threshold value is equal to zero.

38. Server apparatus in a net (n) , which server apparatus is connected to a power network (PN) and a transmission network (IP), wherein at least one measurement is performed of a reference frequency (f_s) in the server apparatus during a predefined time interval, which time interval starts and ends when a phase of a signal, generated by the power network (PN) , equals a threshold value, wherein the server apparatus comprises

measuring means (407] for measuring the reference frequency (f_s) ;

oscillator means (409) for generating the reference frequency (f_s) ;

calculating means (415) for calculating a difference between two measures stored in a memory means (411) , which measures are obtained at the start and at the end of a measurement;

memory means (411) for storing measures performed by the measuring means (407) and results from the calculating means (415);

detector means (405) for detecting when the phase of the signal, generated by the power network PN, equals the threshold value and for calculating the cycles of the reference frequency

( f s i ' and control means (419) for controlling said means .

39. Server apparatus according to claim 38, wherein the server apparatus further comprises

filter means for processing results from a number of measurements; and

comparator means for determining if results of corresponding measurements are different.

40. Server apparatus according to claim 38 or 39, wherein the calculating means (415) calculates a mean value of results from several measurements.

41. Server apparatus according to claim 38, 39 or 40, wherein the mentioned time interval is a number of cycles of the signal, generated by the power network (PN) .

42. Server apparatus according to claim 41, wherein said threshold value is equal to zero.