HK1119854A - Method and apparatus for synchronizing base stations - Google Patents

Description

Method and device for synchronizing base station

The patent application of the invention is a divisional application of an invention patent application with the international application number of PCT/US03/03351, the international application date of 2003, 2, month and 4, the application number of 03803296.1 entering the China national stage and the name of a method and a device for synchronizing base stations.

Technical Field

The present invention relates generally to digital communication systems, and more particularly, to a system and method for synchronizing a plurality of base stations in a cellular communication system.

Background

Proposal 3^rdThe 3G-generation wireless communication protocol requires a simple basis, but requires expensive procedures for externally synchronizing the various base stations to a high-precision external source. Its technique of supporting base station synchronization requires the base station to passively listen fromIts synchronized transmissions on a channel, such as the Synchronization Channel (SCH) or a neighbor of a Common Control Physical Channel (CCPCH), and employs procedures similar to those performed by the User Equipment (UE) for synchronization. Another approach requires each base station to occasionally transmit a particular synchronization burst based on one or more neighbor listens for the transmission. Yet another method has the UEs measuring the time difference of arrival (TDOA) of transmissions from each of the two cells. These techniques utilize a precise source in each base station. These techniques are expensive and inconvenient because each base station has a source of cost.

Therefore, there is a need for a system and method for fast, efficient, and less expensive synchronization that allows for no additional physical resources to be consumed between any base stations.

There is also a need for providing a precisely synchronized system using a minimum number of interruptions for normal service, such as reducing the traffic between the node B and the Radio Network Controller (RNC).

Disclosure of Invention

A method and apparatus for synchronizing base stations using an independent synchronization source, or identifying a base station, such as a master base station, an RNC (C-RNC) or a base station may designate a base station or a UE to request measurements from the base station to accomplish synchronization. Synchronization activity may be ordered periodically or may be granted when periodic measurements indicate that a drift value exceeds a given threshold. The method for synchronizing the base station at least comprises the following steps: a) detecting at least one of a plurality of asynchronous base stations; a1) determining a base station having a better time quality synchronization quality than the at least one asynchronous base station; b) measuring a plurality of cell time signals including cell time signals measured by the better time synchronization quality base station; and c) correcting at least one of the plurality of asynchronous base stations based on the cell time signal measurements and without using timing signal measurements from any base station having worse time synchronization quality than the at least one of the plurality of asynchronous base stations.

The objects and advantages of the preferred embodiment systems and methods will become apparent to those skilled in the art from a reading of the detailed description of the preferred embodiments.

Drawings

Embodiments of the invention will be described with reference to the accompanying drawings, wherein like numerals represent like elements in the drawings, and wherein:

fig. 1 is a block diagram of a communication system.

Fig. 2 is a block diagram of a Radio Network Controller (RNC) according to the preferred embodiment of the present invention.

FIG. 3 is a block diagram of a base station and UE in accordance with the preferred embodiment of the present invention.

FIG. 4 is an illustration of a design of the step time profile generated by the preferred embodiment of the present invention.

Figures 5a and 5b, taken together, comprise a flow chart of a system according to a preferred embodiment of the present invention.

Detailed Description

Fig. 1 illustrates a simplified wireless spread spectrum Code Division Multiple Access (CDMA) or Time Division Duplex (TDD) communication system 18. The system 18 includes a plurality of node Bs 26, 32.. 34, a plurality of RNCs 36, 38.. 40, a plurality of User Equipments (UEs)20, 22, 24, and a core network 46. A node B26 within system 18 communicates with associated user equipment 20-24 (UE). The node-B26 has associated therewith a single base station 30', or a plurality of base stations 30₁...30_nA single position controller (SC) 30. Each base station serves an associated geographic area known as a cell. It should be noted that even though base station synchronization is disclosed, cell synchronization can be accomplished using the present invention.

A group of node Bs 26, 32, 34 are connected to a Radio Network Controller (RNC)36 by an Iub interface. The RNCs 36.. 40 are also connected to the core network 46 by an Iub interface. It is to be understood that only one node B is presented below, but the present invention can be easily applied to a plurality of node Bs.

There are two basic ways in which node B synchronization can be managed-either centralized or decentralized. In the centralized approach, all sub-functions of cell measurement and cell base correction are performed on demand by the controlling RNC (crnc) and measurements are reported to the RNC. In the decentralized approach, there is no direct instruction from the RNC to perform some or all of the node B functions. There is also a different degree of centralization. For example, the preferred method is almost completely centralized, but not allowed for a limited independent function; for example, as discussed below, the node B may independently adjust its internal frequency source based on observed trends in timing corrections sent by the RNC.

An example of a decentralized approach is for the RNC36 to tell each of the cells in the node Bs 26, 32, 34 which cell is used for synchronization and then the RNC allows each of the cells of that cell to adjust its clock signal independently without explicit notification of the RNC's time changes. In the present technique the cells must maintain an accurate clock signal and because all cells are interdependent adjusted, overall system stability may not be assured. The present dispersion method has been proposed, but it is not a preferred method.

According to a preferred technique, the RNC36 maintains synchronization of all base stations within the node Bs 26, 32, 34 and between the node Bs 26, 32, 34. Referring to fig. 2, the RNC36 includes a database 59 having a covariance matrix 57, a synchronization controller 55, a message generator 53 and a measurement receiver 54. The message generator 53 through which the RNC36 can request messages from a base station 30₁...30_nOr measurements of the UE20, 22, 24; the measurement receiver 54 through which the measurements are received; the best-use synchronization controller 55 updates its evaluation based on these evaluation states; and manages a set of states stored in a covariance matrix 57. The stored states are used for synchronization and represent time errors of each base station 30 with respect to a reference,the rate of change of each time error and the time error at the base station 30₁...30_nThe propagation delay therebetween.

The RNC36 also manages a set of measurements stored in a database 59, which includes: the time of arrival of a measured waveform (i.e., synchronization burst); TDOA of transmissions from two base stations as measured by a UE 20; and evaluation of state uncertainty and measurement uncertainty. The database 59 further contains evaluations for all important states, e.g., for all cells (node bs) except the host, time offsets or errors (in nanoseconds, or microseconds, typically values in the range of +/-3 microseconds to +/-3000 nanoseconds); the time rate of change of the time compensation, for example a nanosecond drift per second or a microsecond drift. A state vector is a configured set of all states, e.g., (Δ t (1), Δ t (2),. -, Δ t (n-1),) Having n node Bs, including the host, node B (0), indicating the state vector X ═ X (1), X (2),. X (m) where

m＝2(n-1)＝Δt(1)，Δt(2)，Δt(n-1)，The RNC36 uses advanced filtering, such as a Kalman filter, to evaluate parameters that define the associated clock signal drift, and sublimation parameters such as the precise range between one component and another.

The preferred embodiment uses a one-phase procedure in which the RNC36 configures a time quality to each base station 30₁...30_n. The RNC36 measures the time quality by selecting a base station, such as a timing reference for all other base stations. Configuring all other base stations that are updated based on measurements and applied correctionsA variable time quality. The temporal quality may be an integer (e.g., 0 to 10). A lower quality value includes a better accuracy. As a transformation, the merit value may be a continuous (floating point) variable. The reference base station (master base station) is preferably configured with a quality value of 0. All remaining other base stations are configured with values that are changed and adjusted with respect to the reference base station. To illustrate the present time quality diagonal design, FIG. 4 shows a master base station in which all slave base stations (slave 1, slave 2, slave 3) are configured with their time quality values changed in relation to the master base station. In one embodiment, the time quality from 2 base stations is configured with respect to the value changed from 1 base station, and the time quality from 3 base stations is configured with respect to the value changed from 2 base stations.

An alternative to a full diagonal master/slave configuration is a paired configuration, requiring that the base stations of each pair hear each other shifted by their own frequency close to the frequencies of the other base stations. The relative number of adjustments is defined by a set of unique weights assigned to each base station and stored in the RNC database 59. The procedure for adjusting each base station is the same as disclosed in the preferred embodiment set forth above, except that both "synchronous" and "asynchronous" base stations are adjusted according to the weights configured at the respective base stations. With different weights, a base station can perform a change of angle of the center between full center to full dispersion. However, in many environments, this type of paired clock signal update may not guarantee that the pair of base stations drifts from a continuous clock signal of the other pair of base stations.

In the real phase clock signal configuration, the RNC36, in the normal mode of operation, updates the covariance matrix 57 for the states stored in the RNC database 59 once every predetermined time unit (e.g., once every five seconds or by a time determined by the operator). The diagonal component of the covariance matrix 57 is the estimated change in time error of the individual base stations with respect to the master base station.

When the time error change of a base station exceeds a predetermined threshold, the RNC36 initiates a message to support the time error update of the base station. The update is performed according to one of three ways: first, to indicate a target groupStation measurements from a neighboring base station 30₁，30₂...30_nThe time of arrival (BSTOA) of a synchronization burst at a base station; second, indicating a neighboring base station 30 with better quality₁，30₂...30_nMeasuring the transmitted BSTOA of the destination base station; or third, a UE20 measures that base station and neighboring base stations 30₁，30₂...30_nThe BSTOA of the synchronization burst.

In the first and second methods, the time of arrival of a base station to base station BSTOA transmission is observed. Referring to fig. 3, a transmitting base station 30₁A known transmission type is transmitted at a predetermined time. The present transmission type may be from the base station 30₁A synchronization burst of the synchronization burst generator 62 is transmitted through an isolator 64 before being radiated by an antenna 70. Receiving base station 20₁The transmit waveform is detected, transmitted through an isolator 66 to a measurement device 68, which outputs a large value when the received signal corresponds to the expected signature of the transmit output. If the receiving and transmitting stations 20, 30 are at the same location and have correctly synchronized clock signals, the output of the measurement device 68 will be generated at the same time as the transmitted waveform. However, clock skew and propagation path delay create a time difference.

The transmission path delay is defined as per equation 1:

r/c + x equation 1

Wherein R/c is the distance R between a transmitting unit and a receiving unit divided by the speed of light c. The term x measures the delay for the device. The quantity R/c is typically configured when the base station is very remote. Radio waves are transmitted at the speed of light, approaching 1 foot per nanosecond, or 3 x 10 feet per second⁸And (5) measuring. The purpose of base station synchronization is to configure the base station within 1-3 microseconds. Thus, when base stations are partitioned by distances according to a ranking of 1/2 miles (1km) or more, these distances have a significant effect on delay. However, for a megacent or microcell divided by ten meters, the distance is not important compared to the measurement accuracy of the configuration.

From these considerations, knowledge of the total number of separations (i.e., distances) is important when attempting to synchronize base stations that are far away (more than 1 km). The correct position is not important when attempting to synchronize base stations that are within about 50 meters. After performing measurements of the BSTOA, the known transmission distances stored in the RNC database 59 are configured and account for misalignment differences in time between base stations.

A third method measures the relative time difference of arrival (TDOA) between two transmissions transmitted by two different base stations, as observed by a UE 20. UE20 measures and reports the observed TDOA between transmissions from the two base stations. The RNC36 sends a message to the UE20 to measure the TDOA of the two base stations. Upon receipt of this message, UE20 receives the transmissions of both base stations via its antenna 72 and isolator 66, and measures TDOA using UE measurement receiver 68, and isolator 66 and antenna 72 transmit the measurements to its associated base station.

If the UE location is known (i.e., its distances r1 and r2 to the respective two BSs are known), and the timing of both BSs is correct, TDOA is, as per equation 2:

(r1-r2)/c equation 2

The measurement error from this value will be an indicator of timing misalignment. As those skilled in the art will appreciate, if the distances r1 and r2 are significantly smaller, as would be true for a mega-sized cell, it will not be necessary to know their values. A measurement observing TDOA, such as the time difference of transmission, would be used directly.

Once a method is selected, appropriate messages are sent to either a base station 30₁...30_nOr a UE 22, 24, 20. If so, a message is sent to the base station 30₁...30_nE.g. base station 30₂Telling the base station 30₂That neighbor is monitored and measured. If the message is sent to a UE 22, the UE 22 is instructed to measure that other base station than its own base station.

Referring back to FIG. 2, once the RNC36 is at itHas been stored in the base station 30 in the database 59₁...30_nThe distance between them, which is checked in order to see if there is a base station 30 with better update than it has₂A neighboring base station 30 of a time quality₁. Upon discovery of such a neighboring base station 30₁Initiating a message to the neighboring base station 30₁To obtain data from an "asynchronous" base station 30₂A measurement of (2). Alternatively, the RNC36 may send a message to the "asynchronous" base station 30₂And request acquisition of a neighboring base station 30₁A measurement of (2). For purposes of this embodiment, a required base station, an "asynchronous" base station 30₂Then acquire "synchronization" base station 30₁A measurement and sends a measurement value back to the RNC measurement receiver 54. The RNC measures the forward measurement value of the receiver 54 to the synchronization controller 55, which calculates the measured time of transmission by subtracting the transmission time r/c.

Once the time of transmission is calculated by the RNC synchronization controller 55, the values are compared to the values stored in the RNC database 59. The RNC synchronization controller 55 then calculates the kalman filter gain and updates the state in the covariance matrix 57 using the difference between the calculated and predetermined arrival times and the common gain. If the difference exceeds a certain threshold, to achieve "synchronization" with other base stations 30 under the control of the RNC36₃...30_nThe RNC message generator 53 will then send another message to the "" asynchronous "" base station 30₂To adjust its time base or its reference frequency. The following two problems are noted:

(1) in a preferred embodiment, the RNC sends a message to the node B to adjust its frequency; however, it may be (as in the state of the third generation partnership project, (3GPP) RAN specification) a message that may not be present, and therefore may not use this feature.

(2) In the present concept, without the need for new measurements, the estimated time error can exceed a threshold and trigger a timing correction, i.e. with a highly reliable estimate of the drift rate, the RNC can correctly identify that a node B is exceeding its allowed time offset simply by using the drift rate to infer the time error.

Base station 30₂With the required adjustments and reporting them to the RNC measurement receiver 54. Updating the database in the RNC36, including the destination base station 30₂The modification of the time reference, the change of its time rate (which cannot be applied if it does not already have a frequency adjustment), the updating of its covariance matrix 57 (including, inter alia, the evaluation of RMS time errors and drift errors), and the updating of its temporal quality.

Referring to fig. 4, a base station whose timing is modified based on a comparison with another base station is never configured to be equal to or better than the quality of a base station from which it is a slave. This procedure ensures stability. To illustrate, if a slave 2 base station is modified, only one value of the slave 2 base station that is worse than the value of the slave 1 base station in time quality may be configured. This ensures that the time quality of a base station will not be synchronized to slave base stations of the same or lesser time quality level, which will eventually cause a set of base stations to drift "asynchronous" to the master base station.

As disclosed above, for adjusting the "asynchronous" base station 30₂The transformation method for taking measurements uses a UE20, 22, 24. If the method is selected by the RNC36, a message is sent to the UE 22 to measure the "" asynchronous "" base station 30₂And a 'synchronous' base station 30₁The synchronization burst of (2). Once the measurements are obtained by the UE 22, the measurements are sent to the RNC36 and processed. Similar to the method described above, the measurements are compared to known measurements stored in the RNC database 56 and covariance matrix 57, and an adjusted measurement is sent to the "asynchronous" base station 30₂。

A flow chart of a method according to the preferred embodiment is illustrated in fig. 5a and 5 b. The RNC36 updates the covariance matrix 57 and database 59 once per cell time (step 501). Detecting a base station 30 at the RNC36₂...30_nWhen the time error change exceeds a predetermined threshold (step 502), to update the time error change of the "asynchronous" base station, the RNC36 determines whether to use a Base Station (BS) to measure BSTOA or a UE to measure TDOA (step 503). If the RNC36 determines to measure the BSTOA, it either sends a message to a neighboring base station of the "asynchronous" base station to measure the time of arrival (BSTOA) of the base station or sends a message to the "asynchronous" base station to measure the time of arrival of the neighboring base station (step 504). The appropriate base station takes the requested measurements (step 505) and transmits the measurements to the RNC36 (step 506).

If the RNC36 determines to measure TDOA, the RNC36 sends a message to a UE to measure TDOA for two base stations (step 507a), one of which is an "asynchronous" base station. The UE measures the TDOA of each base station (step 507b) and transmits the measurements to the RNC36 (step 507 c).

Upon receipt by the RNC36 of the appropriate measurements (step 508), the RNC36 compares the measurements to values stored in the RNC database 59 (step 509). If the difference exceeds a specified threshold, the RNC36 sends a message to the "asynchronous" base station to adjust its timing or its reference frequency based on the difference (step 510). The "asynchronous" base station generates the requested adjustment (step 511) and reports it back to the RNC36 (step 512). The RNC database 59 and covariance matrix 57 are then updated to include the new values (step 513).

A preferred embodiment is a system and method that pertains to each RNC 36. In the prior art, a controlling radio network controller (C-RNC) communicates directly with its base stations and a serving radio network controller (S-RNC) communicates directly with its UEs. For states where neighboring base stations are under control of different RNCs, there may be a need to increase communication between C-RNCs and S-RNCs that it controls neighboring base stations and UEs.

Instead of a structure according to a fully centralized control, an alternative embodiment may update the structure according to a more decentralized one. In this embodiment, each member of a pair of base stations that can listen to each other can move its unique frequency closer to the unique frequency of the other member. The relative total number of adjustments is defined by a set of single weights that are allocated to each base station and stored in the RNC database 59. The procedure for adjusting each base station is the same as disclosed in the preferred embodiment above, except that both "synchronous" and "asynchronous" base stations are adjusted based on the weights configured at each base station. With different weights, a base station can perform a change of angle of the center between full center to full dispersion.

In either centralized or decentralized approaches, where multiple cells maintained within a single node B are "synchronized," there are many possibilities. The preferred embodiment enables an RNC36 to transmit time corrections and/or frequency corrections to a base station 30₁...30_n. The master base station is responsive to ensure that each of its base stations within a single node B has a time reference from it that is accurate to within a certain limit. The RNC36, according to its algorithms and modifications, assumes a negligible error between the primary base station and its base station and therefore assumes that all base stations have the same time reference.

Thus, the RNC36 does not attempt to evaluate individual timing errors between the master base station and its slave base stations, and because the associated RNC36 does not perform a correction, the master base station must eliminate or compensate for timing errors between the respective ones of the master base station and the other base stations. This embodiment presents a distinct interface between an RNC36 and a host base station. It enables the master base station to apply its own unique solution to comply with the synchronization that it will be adapted to the megabyte base station. The present approach also reduces the total number of over-the-air synchronizations that it is necessary to measure only one cell of a node B to know the current time and frequency references for all cells within the node B. However, this is a large consequence in the hardware of the node B, since the clock signal reference has to be transferred between the node B location controller (SC) and a cell, and if the distance between SC and a cell is large, the present disadvantage arises only from the time error of the distance.

In a first alternative embodiment, referred to as a "cell primary frequency and time reference", each base station has an independent time and frequency reference that enables an RNC36 to transmit time corrections and/or frequency corrections to each base station. The RNC36 evaluates the state of time and frequency errors that represent each base station according to its algorithms and corrections. Thus, the RNC36 attempts to evaluate the individual time errors between each base station and the main base station, and measurements on one base station do not provide help to evaluate the status of another base station. Thus, the base station manufacturer need only provide a sporadic finite error in the base station's time and frequency drift, and each base station must have an acceptable connection over the air to another base station (the same or a different base station).

This alternative embodiment facilitates large and small sectorized areas where the distance between base stations is remote. The ability to modify a base station's time reference compliance with a node B is limited by measurements with respect to another base station that is compliant with the same node B.

In a second alternative embodiment, referred to as "SC primary frequency reference/cell primary time reference", each base station uses an independent time reference, but the primary base station provides a frequency reference. An RNC36 transmits time corrections and/or a single frequency correction for each base station individually to a master base station. The RNC36 ensures that the clock signal of each base station is compliant in frequency with the clock signal of the main base station. The RNC36, according to its algorithm and modifications, assumes negligible error between the main base station and its configured base stations, but evaluates the compensation for it as a constant. The RNC36 evaluates the time error between the main base station and its base station, and the common frequency drift of the base stations belonging to the main base station.

A third alternative embodiment has similar characteristics to those of the "SC primary frequency and time reference" embodiment, but where the base station is remote from the primary base station. The present embodiment provides a means to remove temporal misalignment in the long range. With the advantage that these time compensations are stable, the present embodiment updates the drift rates for all base stations that are compliant with the same master base station with the advantage of measurement of the frequency of the clock signal that is compliant with any base station to the master base station.

In a fourth alternate embodiment, referred to as "secondary SC primary frequency and clock signal reference", the RNC36 provides the evaluation and the primary base station to support its synchronization subject to its base stations. An RNC36 transmits time corrections and/or frequency corrections for each associated base station to its respective primary base station. The master base station ensures that each of its associated base stations has a time reference subject to itself, accurate to within a certain limit. The master base station may choose to use a single base station evaluation to assist in base station synchronization. The RNC36, based on its algorithms and corrections, establishes an optimal estimate of the time and frequency errors between the primary base station and its base stations. In performing state estimation, the relative confidence in the weights between measurements and base station error uncertainty is estimated. Thus, the RNC36 attempts to evaluate individual time errors between the main base station and its base stations, and the main base station evaluates and/or compensates for time errors between the main base station and the various base stations that are subject to its time references, or requires assistance from the RNC 36. This embodiment allows for a structure similar to the "SC master frequency and time reference" embodiment, but allows for adjustment for less accurate delivery from the master reference, freeing up some of the limitations of that embodiment.

In all timing models, the network is correctly synchronized using a minimum number of interruptions of normal service. This reduces the total number of shadows in the air interface and reduces the number of messages walking on the IUB interface, resulting in a reduction in the total number of headers needed to support node B synchronization as described above.

In High Chip Rate (HCR) TDD and Low Chip Rate (LCR) TDD systems, the use of blanking signals is required for any node B to generate the measurements required by the RNC. HCR TDD systems blanking facilitates the use of a predetermined ordering and typically only requires a node B (for the purpose of measuring its TOA by another node B), transmitting a blanking signal so that a measurement can be generated. LCR TDD systems require the transmitting node B, as some of its neighboring cells, to be blanked to avoid interference generated by the measuring node B by these neighboring cells on the received signal. As those skilled in the art will appreciate, the use of excessive blanking signals in the system interferes with the normal operation of the system, resulting in attenuation.

As disclosed above, the node B synchronization procedure according to the present invention, (either centralized or decentralized), is related to the same function (and sub-function):

1) generating cell measurements

a. The cell transmitting the burst is instructed to transmit the burst.

b. Tells the cells in the vicinity of the transmitting cell to blank their downlink physical synchronization channel (DwPCH) and to generate a measurement.

c. Reporting the measurements as necessary.

2) A cell adjustment is generated that is based on the timing of one or more cells.

It should be noted that equations 1 and 2 may not be synchronized. There may be no one cell adjustment to produce multiple cell measurements and there may be multiple cell adjustments for a single cell measurement.

In the centralized approach, all sub-functions of cell measurement are performed according to the same ordering message, and cell adjustment is requested by the controlling RNC (C-RNC). In the decentralized approach, each sub-function of cell measurement with respect to a detach message and node B can now perform cell adjustment procedures independently.

If a Physical Random Access Channel (PRACH) burst is substituted for DwPCH in the above function, where HCR-equivalent TDD node B synchronization is performed, all UEs in the cell must be aware of the procedure to blank the uplink PRACH slot if it is needed for synchronization, taking advantage of the difference in the status of the usage of the uplink PRACH burst.

All of these messages that require the use of Iub and their traffic load may be a factor.

However, the message instructs a node B to modify its timing impact Iub, but not air interface resources. The decentralized approach uses separate messages, but without a procedure, will result in more messages on the Iub but they are shorter messages. However, Iub load switches are related to more than the message size, so the number of messages is a factor in the Iub load.

To eliminate the total number of blinding generated on the system, from the need for node Bs to generate measurements to ensure continuous synchronization, a fifth alternative embodiment uses the RNC's ability to track long term drifts of individual node Bs with respect to a defined reference. As disclosed above, the RNC may send messages to the node B to generate a measurement, to blank a transmission, and to generate corrections to its timing. The messages may be transmitted according to a predetermined procedure, such as periodically (hourly, per second, etc.). The use of long term drift rates for individual node bs reduces the period necessary to generate measurements. If the short term drift is not a factor, the RNC can maintain synchronization in steady state with a very low rate measurement requirement. Thus, the rate of measurement demand will be directly related to the long term drift.

For example, if node b (a) has a long term drift rate of X minutes per day, the RNC may generate node b (a) a measurement requirement for the total number of times node b (a) has drifted by some time reference more than three microseconds, the time between any pair of cells at which the largest error in the frame begins, e.g., the maximum error in the frame. The total amount of time is based only on the long drift rate. Thus, the periodicity of the measurement requirement will equal its total amount of time for node b (a) to drift by 3 microseconds.

It is only necessary that a given node B can measure the TOA of transmissions of one other node B. As explained above, it can be determined that one of the two node bs has a more accurate time base. The RNC may select one of the two node bs to indicate a modification. In the simplest example, a given time reference measures the TOA from another node B (i.e., slave). The RNC uses measurements to improve the evaluation of the time error and drift of its slaves.

Since the short term order drift (stability) is a factor, the measured rate is driven by the short term order stability, e.g., relative to the long term order stability. In fact, the RNC can get a very accurate estimate of the long term drift rate for a given node B based on past history, but the drift rate can change, thus requiring new measurements. These new measurements are taken when the growth time rate of uncertainty exceeds a threshold. The present time rate value of the growth of uncertainty (maximum allowable error) can be derived from any measurement stored in the RNC. It is known in the art to use stored measurements for the method of determining the present rate. The frequency of the Iub correction message will be symmetric to the long term drift rate and inversely symmetric to the maximum allowable error, which will be higher than the frequency of the over-the-air measurement.

The message suggesting the current set for node B synchronization from the RNC to the node B includes information for the RNC to tell the node B to blank a transmission, generate a synchronization transmission, perform a measurement, or generate a time base correction. Another message is provided instructing a node B to generate an N set of measurements, take an average and then either report the average to the RNC or generate a correction. The instructions may be through a periodic process or as a single event. These new messages may help reduce Iub traffic, but they will not reduce the need for blanking to support measurements.

Some other methods of reducing the Iub message rate include:

1) providing a new message instructing a node B to modify its clock signal rate by n1 ppm; n1 is a predetermined number.

2) Provides a new message instructing a node B to modify its frequency reference (which drives the clock signal) by n1 ppm.

3) A parameter is provided to a node B, which instructs the node B to increase (or decrease) the already existing cell adjustment message of its clock signal every n2 frames by n1 chips, to repeat the adjustment by increasing the number of frames.

4) A requirement for the node B is utilized to obtain the time-corrected drift rate from its RNC and to adjust its clock signal independently.

Methods 1 and 2 require the RNC to send an additional message in the presence cell adjustment message instructing the node B to adjust its clock signal rate or frequency rate to a definite total number. The message is sent at periodic times based on the long term drift rate of the node B. For example, if the RNC determines that the node B clock rate should be adjusted every ten (10) microseconds, a message is sent every ten (10) microseconds, which indicates the total number of adjustments.

Method 3 requires the RNC to send a single message to the node B indicating how long (rate adjusted) to update its clock signal rate based on the estimated long term drift rate calculated by the RNC using measurements stored therein. Since the RNC can calculate the long-term drift rate, it can continuously adjust the long-term drift rate of the node Bs with a single message without Iub traffic, and then the RNC will have to consider only the short-term drifts along with possible changes according to the long-term drift rate, which does not have to repeatedly generate the same correction over time. The message is transmitted only once. The node B continues to update its clock signal rate or frequency reference at the received adjustment rate until the RNC determines that the maximum allowable error has been reached and requests a measurement from the node B from which to adjust its estimated long term drift rate.

Method 3 is also the simplest and can be performed with less extra functionality for timing adjustment messages. Also, this allows the RNC to know the node B behavior (a drawback of the decentralized approach).

Two options for the management of the measurements disclosed above:

the RNC requires measurements when its evaluation non-certainty (according to tracking techniques) exceeds a threshold;

the RNC may simply sort before a measurement sort, assuming an important priority.

The first option best reduces interference with respect to the air interface by ordering measurements only when needed, but increases Iub traffic. The second option reduces Iub traffic. By appropriately configuring the different measurement update rates, the RNC may be (historically determined) adjusted for differences in drift characteristics according to the individual node bs. Either of these two options is an ordering of the strength loss associated with the air interface and Iub requirements over the current LCR node B candidate method, and either option can be performed using the presence message set.

Method 3 is not required but will provide additional reduction in terms of Iub traffic. Its use for LCR TDD can be accomplished by adding a simple modification to the cell synchronization adjustment message.

Method 4 removes some of the tracking algorithms in the RNC and incorporates them in the node B. The node B obtains its drift rate from the time correction from the RNC, and determines an adjustment rate based on its drift rate. The node B adjusts its clock signal according to the determined adjustment rate.

While the invention has been discussed in terms of preferred embodiments, those skilled in the art will recognize other variations that are within the scope of the invention, which is set forth in the claims.

Claims (6)

1. A method for time synchronizing at least one of a plurality of base stations with an independent time reference in a wireless communication system, comprising:

generating a time synchronization message from a covariance matrix database and the independent time reference; and

transmitting the time synchronization message to at least one of the plurality of base stations, wherein the time synchronization message is not based on time synchronization information from any base station having a worse time synchronization quality than the at least one of the plurality of base stations.

2. A method for time synchronizing at least one of a plurality of base stations with an independent frequency reference in a wireless communication system, comprising:

generating a frequency synchronization message from a covariance matrix database and the independent frequency reference; and

transmitting the frequency synchronization message to at least one of the plurality of base stations, wherein the frequency synchronization message is not based on frequency synchronization information from any base station having a worse frequency synchronization quality than the at least one of the plurality of base stations.

3. A Base Station (BS) for a wireless digital system having the capability to time synchronize a plurality of other base stations to an independent time reference, the base station comprising:

means for generating a time synchronization message; and

means for transmitting the time synchronization message to a plurality of other base stations, whereby the received timing of the time synchronization message is used to update a time synchronization timing from only a base station having a better time synchronization quality than the base station.

4. A Base Station (BS) for a wireless digital system having the capability of frequency synchronizing a plurality of other base stations to an independent time reference, the base station comprising:

means for generating a frequency synchronization message; and

means for transmitting the frequency synchronization message to a plurality of other base stations, whereby the received timing of the frequency synchronization message is used to update a time synchronization timing from only a base station having a better time synchronization quality than the base station.

5. A method for synchronizing base stations with a predetermined time reference time in a wireless communication system having an RNC, comprising:

generating a time synchronization message from the database of the RNC; and

transmitting the time synchronization message to at least one of the plurality of base stations, the time synchronization message not being based on measurements from any base station having a worse time synchronization quality than the at least one of the plurality of base stations.

6. A method for synchronizing base stations with a predetermined frequency reference frequency in a wireless communication system having an RNC, comprising:

generating a frequency synchronization message from the database of the RNC; and

transmitting the time synchronization message to at least one of the plurality of base stations, and the frequency synchronization message is not based on measurements from any base station having a worse frequency synchronization quality than the at least one of the plurality of base stations.