AU700042B2

AU700042B2 - Cellular mobile radio system synchronization

Info

Publication number: AU700042B2
Application number: AU30134/95A
Authority: AU
Inventors: Pierre Dupuy
Original assignee: Alcatel NV
Current assignee: Alcatel Lucent NV
Priority date: 1994-09-05
Filing date: 1995-08-21
Publication date: 1998-12-17
Anticipated expiration: 2015-08-21
Also published as: EP0700228A1; FR2724281B1; AU3013495A; NZ272807A; EP0700228B1; FI954106A0; DE69524692D1; FI954106A; DE69524692T2; ATE211342T1; FR2724281A1

Abstract

The method involves calculating the difference (RTD=R12-0, R12-1) between a first time reference of a first oscillator and a second time reference of a second oscillator at two successive instants (t0,t4). This is determined as a function of the propagation time difference associated with the first and second oscillators. The variation between the calculated reference time differences is determined by a time unit. The frequency difference between the oscillators is determined in response to a variation in the calculated time differences which is greater than a given threshold.

Description

P/00/011 28/5/91 Regulation 3.2

AUSTRALIA

Patents Act 1990

ORIGINAL

COMPLETE SPECIFICATION STANDARD PATENT Invention Title: "CELLULAR MOBILE RADIO SYSTEM SYNCHRONIZATION" The following statement is a full description of this invention, including the best method of performing it known to us:- This invention relates to maintenance operations connected with base station oscillators in a radiocommunications cellular network with mobiles, and in particular to a method of detecting the frequency drift of a base station oscillator and which aims to optimise the maintenance operations.

In a base station of a digital radiocommunications network with mobiles, digital transmission over a radio channel is organised in frames. In the context of G.S.M.

(Global System For Mobile Communications), each frame has a total duration of 120/26 ms, that is 4.615 ms and includes 8 time intervals or time slots of 0.557 ms defining 156.25 time-bits. These frames are combined into superframes each comprising 1326 frames and defining a sequence of 6.12 s. The superframes are combined into a hyperframe of 2,048 superframes, that is, a sequence of 3h 28mn 3 s 760ms.

A base station uses these various frame, superframe and hyperframe times to progress program sequences. For example, the frame defines the elementary timing pattern of channel transmission, and the hyperframe corresponds to the coding cycle.

An oscillator in the base station is designed to produce a time reference clock signal for the various operations of frame slot opening, timing cycle of the frequency jumping distribution, etc... Provision is made for the clock signal to have a stability of 0.05 ppm (parts per million), that is to say, a maximum frequency variation of (0.05/10 6

).F

•o~ around the nominal frequency F of the oscillator.

toIn accordance with the prior art, periodic preventive maintenance operations ~are designed to check that the frequency drift of an oscillator either side of the nominal frequency has not brought the oscillator's frequency outside the range [F (0.05/10 6 F (0.05/10 6 These preventive maintenance operations are o t o" expensive since they necessitate on-site intervention by a skilled worker. Furthermore, because of their periodicity, they do not allow the earliest possible detection of frequency drift of an oscillator.

The invention aims to remedy the above-mentioned drawbacks in the prior art by providing a method of detecting the frequency drift of an oscillator in a base station. The method eliminates any preventive maintenance operation by restricting the role of workers to standard single unavoidable adjustment or exchange action.

The resulting economic benefit is obvious. Moreover, detection of the frequency drift of an oscillator is more or less immediate, and action is not deferred until the next preventive service on the oscillator by a worker. Consequently, excessive drift of an oscillator is avoided since such a drift is immediately detected.

According to the invention there is provided a method of detecting the frequency drift between a first oscillator and a second oscillator associated respectively with a first radio transmitter and a second radio transmitter, including the following steps: calculating the difference between a first time reference of the first oscillator and a second time reference of the second oscillator, for each of at least two successive instants, as a function firstly of a difference observed by a relay between the first time reference and second time reference respectively transmitted by the first and second transmitters, and secondly of the respective propagation times between the said relay and the first and second transmitters, the said first and second transmitters together forming a first pair of transmitters, determining the variation per unit time in the calculated time reference difference, and detecting the frequency drift between the said first and second oscillators, in response to a variation per unit time in the calculated time reference difference which is greater than a predetermined threshold.

Preferably, the method may include a selection step for selecting one of two oscillators which have frequency drift as a function of the variation per unit time in the time reference difference determined for at least one other pair of base stations 126: distinct from the said first pair and including one of the said first and second transmitters.

Preferably, the first and second radio transmitters are respectively first and second base stations in a cellular radiocommunications network with mobiles, and the relay is a mobile in the said network.

Accordingly, provision is made for the transmission, by the said mobile to the second base station during a cell handover, of data having a value which is the result of summing, firstly, the difference in time references between first and second base stations as observed by the mobile, and secondly, the propagation time between the mobile and the first base station. Moreover, the step calculating the difference in time references consists in subtracting the propagating time between the mobile and the second base station from the said data transmitted by the mobile to the second base station.

The invention also provides infrastructure equipment for implementing the invention. It includes means for receiving messages from a pair of base stations, each message including a calculated time reference difference value, means for storing for this pair of stations the said calculated time reference difference values at successive instants, and processing means for determining the variation per unit time in the time reference difference, and for detecting the frequency drift of an oscillator in response to a variation greater than a predetermined threshold. The equipment can include, furthermore, means for selecting one of the two oscillators which has frequency drift .1.4 as a function of the variation per unit time in the difference in time references which is determined for at least one other pair of base stations distinct from the said first pair and including one of the said first and second transmitters.

In order that the invention may be readily carried into effect, embodiments thereof will now be described in relation to the accompanying drawings, in which: *2 C Figure 1 illustrates schematically a cell handclover of a mobile in a cellular radiocommunications network with mobiles, and message transfers between mobile and source base and target base stations involved in the cell handclover; and Figures 2 and 3 show respectively a time diagram of instants of cell handover, and an associated Cartesian diagram representing the development in the course of time of three time reference differences associated respectively with three pairs of base stations, in order to explain the implementation of the method according to the invention.

The invention is particularly intended to be implemented within the G.S.M.

which envisages, with reference to Annex A of GSM Recommendation 05.10, Version 4.0.0 (February 1992), published by ETSI, that a mobile plays the role of a "relay" transmitting synchronisation data between two base stations, called "source base station" and "target base station", during a cell handover of the said mobile.

The relay function of transmitting synchronisation data between source and target base stations which the mobile puts into effect is now explained with reference to Figure 1. When a mobile M crosses an imaginary boundary F separating two adjacent cells in a radiocommunications cellular network with mobiles, a cell handover procedure is activated and consists in switching the call established with the mobile M from a first duplex channel associated with the source base station BSO to a second duplex channel associated with the target base station BS1. The cell handover procedure is characterised by predetermined message transfers between the source base station or target base station on the one hand, and the mobile on the other. This oo :i procedure furthermore consists in establishing synchronisation of the mobile with the target base station BS1 whose time reference differs from that of the source station BSO with which the mobile is synchronised. The cell handover procedure for a mobile and the steps pertaining to it are described in French Patent Application FR-A-0 398 .21 773.

0 Attention will be focused solely on a final phase of the cell handover which is helpful in understanding the invention. The difference in time references as seen by the mobile, marked OTD for (Observed Time Difference), between the source base 0090 station BSO and target base station BS1, is defined by: OTD RTD T1 TO 9.

(1) where RTD designates the actual difference in time references (Real Time Difference) between the source and target base stations, or time reference difference, and TO and T1 denote respectively the propagation times between the mobile M and the source base station BSO, and the mobile M and the target base station BS1.

Periodically, typically at a frequency of twice per second in the GSM, the mobile M receives from the source base station BSO the time advance data TO which authorises a time advance adjustment to the clock of the mobile compared to a first time reference which it has received from the station BSO on establishing the call, by taking account of the mobile source base station propagation time. The first time reference is conveyed in the subchannel SCH (Synchronisation CHannel) of the channel BCCH (BroadCast CHannel) in the GSM in the form of a message defining the rank in the hyperframe of the frame which is conveying the said message.

Furthermore, when the mobile M is in communication with the said source base station BSO, it is pre-synchronised with each neighbouring base station to which it is likely to be transferred by handover. This pre-synchronisation is obtained by the mobile's reception of a second time reference transmitted by each neighbouring station in a subchannel SCH. The mobile M is therefore capable of calculating the value (OTD TO). OTD is defined as the difference in the first and second time references, as received by the mobile M with respective propagation delays.

o Recommendation GSM-05.10, to which reference has previously been made, specifies that the mobile transmits this data (OTD TO) to the target base station S BS1 at the end of a handover procedure in a message COMPLETE". In addition, Sduring the handover procedure a message ACCESS" is transmitted by the mobile M to the target station BS1 in synchronism with the second time reference 206 received from the target base station BS1. In response to the reception of the "H.O.

ACCESS" message by the target station BS1, the latter calculates the time advance, or propagation time between the mobile M and itself, which it transmits to the mobile M in a message "PHYSICAL INFO", so that the mobile adjusts its clock in relation to the 0 time reference of the station BS1, by taking account of the mobile station BS1 propagation time T1, and can validly open receiving and transmitting time slots synchronised with the station BS1.

Thus, at the end of the cell handover procedure, the target base station BS1 is in possession of the data (OTD TO) and T1. In accordance with Equation given above, these data (OTD TO) and T1 allow calculation of the time reference difference between the source base station BSO and the target base station BS1 involved in the handover.

With reference to Figures 2 and 3, implementation of the method according to the invention is now explained. The method uses to advantage the characteristic according to which a base station, which operates as a target base station in a cell handover, uses the values (OTD TO) and T1 which are necessary for calculating the time reference difference RTD between itself and a source base station BSO.

In Figure 2, six cell handovers, associated with base stations respectively marked 1,2, 3 and 4, are respectively associated with at six successive instants marked tO, tl, t2, t3, t4 and t5, each shown on the time axis t. On the left in Figure 2, the annotation associated with the source base station is marked i, and the target base station is marked j. R designates the real timereference difference (RTD) between the source base station and the target base station, calculated by the target station for each cell handover. Thus, by way of example, at instant t2 a cell handover is carried out from the cell associated with station 1 to the cell associated with station 4, and the calculated time reference difference is marked R1 40 The specialist will appreciate 1 5 that the calculation of the real time reference difference RTD is symmetrical, which signifies that at a given instant, two identical calculations are obtained if the respective source and target roles, played by each of the two base stations involved in the *i handover, are reversed.

.In Figure 3, in connection with Figure 2, the development over time t of the real F265: time references difference RTD is shown for each of three pairs 3) and of base stations between which cell handovers take place. A first analysis of the results X schematised in Figure 3 establishes that one of the two base stations 1 and 2 has an oscillator which has frequency drift relative to the oscillator in the other of the said stations. This analysis consists, for each pair of base stations, in determining the eeo variation per unit time in the real time reference difference between the two stations.

This variation per unit time is written (R 121 R1 20) (t4 tO) for the pair of S: stations 2) for the instants tO and t4. In Figure 3, for example, an oscillator in one of the stations of the pair stations 2) has frequency drift. It is impossible at this stage to know which of the two stations 1 or 2 has the frequency drift. Nevertheless, a more thorough analysis leads us to find that only station 1 has a frequency drift compared to the other stations. In effect, only the base stations 1 and 4 in the pair of stations 4) show equally a frequency drift between themselves. In contrast, stations 2 and 3 in the pair 3) show no frequency drift between themselves. Station 1 belongs to the pairs 2) and 4).

As a result of what has been written previously, two variants can be designed according to the invention in order to establish the frequency drift of an oscillator in a base station.

According to a first variant, detection of an oscillator with frequency drift is obtained separately for each pair of first and second base stations j) between which cell handovers take place. For each one of at least two successive instants and as a function, firstly, of a difference OTD observed by a mobile between the first and second time references respectively transmitted by the first and second base stations, and, secondly, respective propagation times between the said mobile M and the first and second stations, this variant includes a first step of calculating the difference between a first time reference of a first oscillator in the first base station and a second time reference of a second oscillator in the second base station. The information OTD is transmitted by two a priori distinct mobiles M at the time of two successive handovers between the stations i and j at the said two instants. This calculation step is followed by a step of determining the variation per unit time in this time reference difference. A variation per unit time in the time reference difference, which is greater than a predetermined threshold, allows detection of the fact that one of the said two respective oscillators of the said first and second base stations has too high a frequency drift, an uncertainty which resides in the selection of one of the two oscillators which has the frequency drift.

The second variant of the invention removes this uncertainty. In addition to the above-mentioned steps, it provides to this end a selection step for selecting one of the two oscillators of the pair of base stations j) which has the frequency drift, as a function of the variation per unit time in the time reference difference which is determined for at least one pair of base stations distinct from the pair in question and including one of the said first and second base stations. This selection step aims to remove the uncertainty, that is to say, which of the two stations of the pair j) has an oscillator requiring adjustment. Thus, by way of example, if the base stations i and j' in a pair of stations show a frequency drift between them, then the station i is selected as having an oscillator for which adjustment action has to be provided.

Similarly, if the stations i' and j in a pair j) show no frequency drift between them or a lesser drift, the station i is selected as having an oscillator requiring adjustment action since the station i shows a priori no frequency drift.

According to a more efficient variant, it will be convenient to examine the pair of stations j) and the pair of stations j) at the same time. Therefore, if the base stations i and j' in a pair of stations show a frequency drift between them, and if the stations i' and j in the pair j) show no frequency drift between them or a lesser drift, the station i is selected as having an oscillator requiring adjustment action.

In the mobile network, or infrastructure, it is possible to provide infrastructure equipment for implementing the method of the invention.

The infrastructure equipment then includes means for receiving from a pair of base stations, operating as target base stations, messages each of which includes a calculated value RTD of time reference difference between source and target base stations, and the pair j) of base and target source stations associated with this calculated value RTD. The equipment also includes a memory designed to store a set of calculated values of time reference differences between stations for each pair of cells, or base stations, between which handovers occur. Furthermore, processing 2: means are provided in order to determine the variation per unit time in the time reference difference for each pair of base stations, and to detect the frequency drift of an oscillator in response to a variation greater than a predetermined threshold. For implementation of a second variant of the method according to the invention, the processing means select one of the two oscillators which shows frequency drift as a function of the variation per unit time in the time references difference determined for at least two pairs of cells of base stations, each of which includes the respective one of the base stations of each of the said pair.

The specialist will acknowledge that the invention is not limited to a cellular radiocommunications network with mobiles, and that in the previous description radio transmitters, which include respective oscillators, can replace the base stations, and the mobile can be any transmission relay.

The claims defining the invention are as follows: 1. A method of detecting the frequency drift between a first oscillator and a second oscillator respectively associated with a first base station radio transmitter and a second base station radio transmitter in a celular radiocommunications network with mobile stations, including the following steps: calculating the difference between a first time reference of the first oscillator and a second time reference of the second oscillator, for each of at least two successive instants, as a function firstly of a difference observed by a mobile station between the first time reference and second time reference respectively transmitted by the first and second transmitters, and secondly of the respective propagation times between the said mobile station and the first and second transmitters, the said first and second transmitters together forming a first pair of transmitters, determining the variation per unit time in the calculated time reference, and detecting the frequency drift between the said first and second oscillators, in response to a variation per unit time in the calculated time reference difference which is greater than a predetermined threshold.

ooo* 2. A method as claimed in Claim 1, including the step of selecting of one of the

S

two oscillators, the selected oscillator having frequency drift as a function of the variation per unit time in the time reference difference determined for at least one other pair of base stations.

3. A method as claimed in Claim 2, wherein said mobile station transmitting data to the second base station on handover between cells, which data has a value that is S" the result of summing, firstly the time reference difference between the first and second base stations, as observed by the mobile station, and secondly the propagation time between the mobile station and the first base station.

4. A method as claimed in Claim 3, wherein the said step of calculating the time reference difference consists in subtracting the propagation time between the mobile station and the second base station from the said data transmitted by the mobile station to the second base station.

5. Infrastructure equipment for implementing the method as claimed in Claim 1, including:

-V

Claims

6. Infrastructure equipment as claimed in Claim 5, for implementing the method as claimed in Claim 2, further including means for selecting one of the two oscillators which shows frequency drift as a function of the variation per unit time in the time reference difference which is determined for at least said one other pair of base stations.
7. A method substantially as herein described with reference to Figures 1 3 of the accompanying drawings.
8. Infrastructure equipment, substantially as herein described with reference to S Figures 1 3 of the accompanying drawings. o° DATED THIS TWENTY-SIX DAY OF OCTOBER 1998 ::AA 20 ALCATEL N.V. o* a. a o. 3QI 9 1/2 FIG.1 0* B B B. B B B B B B B B. B B B B B SO BS1