A METHOD FOR SYNCHRONIZATION OF A CELLULAR RADIO NETWORK
FIELD OF THE INVENTION
The present invention relates to a method for synchronization of a cellular radio network. In particular, the present invention concerns the synchronization of base stations in a base station system including a plurality of base stations and a base station controller, of a cellular radio telecommunication network which operates, for example, according to the GSM standard.
BACKGROUND OF THE INVENTION
Fig. 1A illustrates an example of such a cellular radio telecommunication network which is briefly explained below. Only for simplification of explanation, the illustration is limited to two stationary radio transceiver devices or base stations BS (BS1, BS2)
(hereinafter also referred to as a first type of radio transceiver) and to a single mobile radio transceiver device or mobile station MS (hereinafter also referred to as a second type of radio transceiver) .
The base stations BS1, BS2 are controlled by control commands transmitted from an associated controller device such as a base station controller BSC. In a cellular radio communication system, each base station BS defines a cell by the radio coverage area, and serves the mobile stations MS present within the cell. The mobile stations MS move (roam) within a respective cell and from one cell to another cell. In case the mobile station moves to a neighboring cell, a process of hand-over will take place, during which the moving mobile station MS will terminate
communication with the base station BS of the cell being left and initiate communication with the base station BS of the cell to which it moves.
Communication between a respective base station BS and the mobile station MS takes place via the air interface Um. The data transmission via the air interface, according to GSM, is effected using a time divisional multiple access method (TDMA) . According to such a method, as schematically illustrated in Fig. IB, the available transmission time on a physical channel (or freguency) is separated into time slots TS. Each subscriber such as a mobile station MS may communicate with a base station BS only during a specified or predetermined time slot TS.
Within one time slot TS, all kinds of data are transmitted as bursts. Each burst has a length of 148 bit transmitted within 546.12μs. The first and the last bits (3 bits) named tail bits T are always set to "0" in order to avoid interference with a neighboring burst. A burst is further followed by a guard period GP of 8.25 bits duration, thereby completing the period of a respective time slot TS (4.615/8 ms). Further, according to GSM specification, eight time slots TS are defined to constitute one frame F, and frames are grouped to form a frame structure of so-called multi-frames.
In order that for a moving mobile station MS it is known to which base station BS communication has to be established, each respective base station BS informs the mobile station via the broadcast control channel BCCH of parameters which are specific for the network.
Moreover, in a standard cell configuration of such a network, each cell is surrounded by a plurality of cells, of which plurality of cells those six neighboring cells which are strongest (in terms of received signal level) are monitored. Thus, since a mobile station within a specific cell may move to an arbitrary one of these six neighboring cells, the mobile station MS monitors up to six neighboring base stations BS by listening to the BCCH channel of the respective neighboring base stations.
The cellular radio communication networks as described herein above are normally not synchronized, since up to now, no low cost mechanism has been available to synchronize the network.
Nevertheless, synchronizing base stations with respect to each other recently receives increasing attention. In particular in connection with a cellular network operating according to TDMA/TDD (Time Division Duplex) , synchronization of base stations is inevitable. That is, if the timing of frames differs from base station to base station, time slots will overlap. In consequence, severe interference problems will manifest. Hence, those overlapped time slots cannot be used by either base station.
Hitherto known approaches to synchronize base stations with respect to each other are for example described in "Introduction to Digital Mobile Communication", by Y. Akaiwa, pp. 390-394, John Wiley & Sons Inc., New York, USA, 1997.
As discussed therein, a first approach resides in broadcasting a reference timing signal. The reference
- A - timing signal may be broadcast based on a signal provided by the GPS system (Global Positioning System) . However, this approach has a drawback that each base station requires a GPS receiver. This, in turn, significantly increases the costs for the hardware of a base station. Moreover, this approach is not fully satisfactory in connection with indoor base stations.
Furthermore, another approach has been proposed, according to which each base station measures the timing errors between its own frame timing and the frame timing of neighboring base stations. The synchronization is established on the basis of an average value of the measured timing errors.
However, this approach suffers from a drawback that additional measuring efforts are necessary to be carried out by each base station. Furthermore, each base station has to be provided with the necessary measurement equipment. Therefore, also this approach is not preferable under consideration of the resulting costs.
SUMMARY OF THE INVENTION
Hence, it is an object of the present invention to provide a method for synchronization of a cellular radio network, which does not rely on additional hardware, and which can be realized at low costs.
In order to accomplish the above object, the present invention proposes a method for synchronization of a cellular radio network, the network comprising at least two radio transceiver devices of a first type, and at least one radio transceiver device of a second type, the method comprising the steps of measuring timing offsets
between respective pairs of said first type radio transceiver devices by means of at least one of said second type radio transceiver devices, controlling said first type transceiver devices by means of a controller device on the basis of said measured timing offsets such that one of said first type transceiver devices is set as a master transceiver device with the transmission and reception timing thereof remaining unchanged, and the remaining first type transceiver devices are set as slave transceiver devices with the transmission and reception timing thereof being adjusted so as to be adapted to the timing of said master transceiver device, with the timing being adjusted such that subsequently measured timing offsets between said master transceiver device and each respective slave transceiver device are minimized.
Thus, the present invention is particularly advantageous in that the proposed method does not rely on additional hardware to be provided. Instead, the equipment provided and measurement results obtained for determination of the position (and speed) of mobile stations are used for synchronization purposes. Therefore, the proposed method represents a low cost method for synchronization of base stations .
Moreover, based on synchronization between base stations, the performance of a cellular radio network, in particular of the radio interface, may be enhanced.
Based on inter-base-station synchronization, orthogonal frequency hopping becomes possible and is improved, and very small frequency reuse factors can be used in a cellular network. Thereby, the traffic capacity of the network is increased. This is particularly beneficial in
densely populated areas requiring a large radio traffic capacity.
Additionally, hand-over of a mobile station between neighboring base stations can be synchronized, thereby improving the transmission quality in the hand-over and the comfort to the user/subscriber of the mobile station, while also the hand-over phase is reduced and the new base station accesses faster.
Furthermore, due to synchronized base stations, interference cancellation becomes possible. This improves the network characteristics in terms of an improved channel interference ratio (C/I or CIR) and also enhances spectral efficiency (efficiency of use of available frequencies) for both, mobile stations and base stations.
Advantageous further developments of the present invention are as set out in the dependent claims.
Preferred embodiments of the present invention are described herein below in detail by way of example with reference to the accompanying drawings .
BRIEF DESCRIPTION OF THE DRAWINGS
Fig. 1A shows schematically a part of a cellular radio telecommunication network;
Fig. IB shows schematically the hierarchical organization of data in a TDMA system;
Fig. 2 is a flowchart illustrating the method for synchronizing a radio network, i.e. base stations, based
on time difference measurements effected by a mobile station;
DESCRIPTION OF PREFERRED EMBODIMENTS
According to the present invention, the proposed method for synchronization of base stations uses timing offsets, in particular observed timing differences (OTD) . The OTD's are observed by each respective mobile station MS with respect to base stations, or a respective pair of base stations BS1, BS2. The value of the observed timing offsets is derived from the information transmitted by the base stations on the BCCH broadcast control channel. Based on the values of OTD, a respective base station is set as a master base station M_BS, and the remaining base stations in a given area are set to be slave base stations S_BS . The base station controller BSC commands the slave base stations S_BS to adjust their timing to the timing of the master base station M_BS such that upon a subsequent measurement of the OTD, the slave base stations operate in synchronism to the master base station. The expression "synchronized" means that each burst sent by one base station does not overlap over two time slots of a neighboring base station.
For better understanding of the present invention, the measurement of the observed timing difference OTD by each of a plurality of mobile stations MS within a cellular network is explained with reference to Fig. 1A.
In Fig. 1A, base station BS1 sends information on the BCCH channel to the mobile station MS at a time tlS. The information is received by the mobile station MS at time tlR. Similarly, information sent from base station BS2 at time t2S is received at the mobile station MS at time
t2R. The offset between the sending times tlS, t2S is the real time difference RTD between the base stations, and hence an indication for asynchronism there between.
RTD = tlS - t2S ... (1)
In particular, it will be apparent that RTD = 0 means that the base stations BSl, BS2 transmit synchronously with respect to each other.
The observed timing difference OTD as measured by the mobile station MS corresponds to the difference in reception timings tlR, t2R. That is
OTD = tlR - t2R ... (2)
The time of reception, in turn, depends on the distance between the mobile station MS and each respective base station BSl, BS2. Stated in other words, the propagation times
tlP = tlR - tlS t2P = t2R - t2S ... (3)
depend on the distances dl, d2 of the mobile station MS from respective base stations BSl, BS2. Thus, a geometric time difference GTD may be defined as
GTD = tlP - t2P = tlR - tlS - (t2R - t2S)
= (dl - d2) /c ... (4)
with c being the propagation speed of the radio waves which may be assumed to be the velocity of light in case of the air interface.
Substituting equation (4) in equation (2), and then substituting equation (1) in the thus obtained equation yields
OTD = GTD + RTD ... (5)
Therefore, assuming a situation in which a mobile station is placed symmetrically between two base stations (i.e. dl - d2 = 0) , equation (4) yields zero for the geometric time difference GTD. Thus, the observed time difference OTD directly corresponds to the real time difference RTD as an indication for the asynchronism between base stations. In any case, from the observed time difference OTD it can be concluded which timing offset between base stations exists.
For example, according to a preferred implementation of the present invention, each mobile station MS determines the observed timing difference OTD in intervals which correspond to a quarter of the duration for the transmission of a bit. According to GSM, one burst contains 148 bit per 546.12μs. Thus, one bit requires 3.69μs for transmission. Therefore, a quarter bit duration corresponds to less than lμs, and OTD values are detected or measured, respectively, nearly every microsecond.
Although this is a preferred implementation, this interval is not strict for the function of the present invention. The interval merely has to be chosen such that between subsequent measurements of the OTD value it can be surely assumed that the geometric time difference GTD remains substantially uninfluenced due to a movement (moving speed) of the mobile station occurring in this period. In such a case, the difference in consecutively
measured OTD values corresponds to the difference in the RTD values. Hence, based on equation (5), it can be concluded that the asynchronism is minimal in case the OTD value becomes a minimum value. This means, that RTD is minimal or even zero. Stated in other words, upon detection of a minimal OTD value, base stations are synchronized with respect to each other.
The method according to the present invention will be described in further detail herein below with reference to Fig. 2. Fig. 2 shows a flowchart which illustrates the method for synchronizing a cellular radio network, i.e., for synchronizing the base stations (first type radio transceiver devices) of the network based on time difference measurements effected by mobile stations (second type radio transceiver devices).
Hitherto, the description focused on a simplified example for the case of only two base stations and one mobile station being present in the network, in order to explain the principle underlying the present invention. In actual networks, however, as stated above, each base station is surrounded by six (monitored) neighboring base stations (or even more surrounding base stations) and in each cell of the network, more than one mobile station may be present. In such a case, the method works as described in the following.
Upon starting the synchronization of base stations BS (step S200) , all mobile stations MS measure the observed timing differences OTD of the respective base stations or of respective pairs of base stations (step S201) . These values are measured to obtain the position of the mobile stations within the network, as for example described in same applicants former application PCT/EP97/02400.
Subsequently, at least one of the mobile stations MS is selected (step S202) , so that the synchronization will be based on the measurement results obtained by the selected mobile station or stations, respectively. The method for selecting the mobile station will be described further below.
In step S203, the OTD values measured by the selected mobile station or stations, respectively, are transmitted via the base station BS of the cell in which the at least one selected mobile station moves, to the base station controller BSC. Upon evaluating the received OTD values, the base station controller BSC, in step S204, sets one base station BS as a master base station M_BS, while the remaining base stations are set as slave base stations S_BS. The method for selecting the master base station will also be described further below.
Subsequently, the base station controller BSC commands (step S205) each slave base station S_BS to change its timing to thereby adapt the timing to the timing of the master base station M_BS . After the instructed change of the timing of the slave base stations S_BS, the mobile stations obtain and transmit the resulting new OTD values to the base station controller BSC (step S206) . The base station controller (step S207) compares the obtained new values of OTD between the master base station M_BS and the respective slave base stations S_BS with the previously obtained values.
If it is judged by the base station controller BSC in step S208 that the respective OTD values are minimal, it is determined that the base stations are synchronized (step S209) and the process is terminated (step S210) . However, if it is judged in step S208 that respective
ones of the OTD values are not minimal, the process returns to step S205. Then, the respective slave base stations are again commanded to adapt there timing to the timing of the master base station. This loop is continued until the slave base stations S_BS are synchronized to the selected master base station M_BS, i.e. until an increase in timing difference is observed between two consecutive measurements. In case an increase is observed, the base station controller BSC sets the immediately preceding timing conditions for the respective slave base stations S_BS, for which the minimum OTD values were obtained, in order to have the network synchronized. Then, the process is terminated.
In the following, the selection of the at least one mobile station MS, performed in step S202, is described in further detail.
Generally, a plurality of mobile stations MS are present within a cell of a radio network. Each of these mobile stations MS monitors the six (strongest) neighboring base stations as well as the base station of the current cell in which the mobile station moves. Thus, a plurality of measurement results of OTD values are present, which are also used for determining the position (and/or speed) of each mobile station within the network. Moreover, based on the determined position of a mobile station, the base station communicating with the respective mobile station determines the timing advance TA for each respective mobile station in the cell in order that propagation delays of signals between base station and mobile stations are compensated for. The timing advance TA is indicated to each mobile station as number of bits. These number of bits corresponds to the advanced timing of
mobile station transmission in order that communication with the base station of the cell may be possible.
However, for network synchronization purposes it is required to rely on OTD values of at least a single mobile station to be selected among the plurality of mobile stations. To this end, it is conceivable that only those measured timing offsets OTD which are obtained by a respective mobile radio transceiver device MS having a timing advance TA below a predetermined threshold value is transmitted to said controller device BSC. Further preferably, said predetermined threshold value is set depending on the geographical environment and/or the density of base stations in the specific area. Stated in other words, only OTD values of those mobile stations may be selected as a basis for control, which meet predetermined conditions for location and speed within the cell (or network) .
In this way, the distance variation of mobile stations to the base station can be taken into account when selecting the mobile station (s) the measurement results of which are used for synchronization.
Moreover, it is preferable to base the control on the OTD values of more than one mobile station, since this will provide better accuracy for synchronization. Stated in other words, measured timing offsets from a plurality of mobile transceiver devices MS may be combined and used as a basis for control.
In this connection it is conceivable to derive an average OTD value from the OTD values observed by respective mobile stations. Apart from a mere average value, also a weighted average value may be calculated. In order to
obtain a weighted average value, each OTD value is weighted by multiplying it with a predetermined factor. Such a weighting factor may for example be set in accordance with the S/N ratio, i.e. quality, of the signal transmission between the respective mobile station and the respective base station. This will lead to a more reliable value for OTD upon which synchronization control is based.
Generally, it is rather likely that a plurality of mobile stations meet one of the above mentioned requirements for being selected. Thus, averaging of OTD values is also applicable in connection with a situation in which a plurality of mobile stations MS fulfills the above described threshold-requirement (and/or location-speed- requirement) . Then, only those OTD values will be averaged, which are obtained by those mobile stations which fulfill the above described threshold-requirement, thereby leading to further improvement in control reliability.
Furthermore, once the OTD measurement results of the mobile station MS are transmitted via the base station BS to the base station controller BSC, the question arises as to which of the plurality of monitored base stations BS is to be selected as the master base station M_BS (step S204 in Fig. 2). This selection can be based on different approaches.
Firstly, the base station BS of the cell in which the selected mobile station (s) MS is (are) present can be fixedly defined as the master base station M_BS . Alternatively, an arbitrary (predetermined) one of the neighboring six base stations may be set as the master base station. In particular, the base station among the
monitored base stations having the strongest signal level in communication with the mobile station (s) MS may be selected as the master base station M_BS . Moreover, the selection of the master base station may be dependent on the measured OTD values. That is, each of the at least one selected mobile station monitors the base station it is currently assigned to (present in the cell defined by the base station) and the six strongest surrounding base stations. Thus, seven base stations are monitored and twenty one different pairs of OTD values between respective pairs of base stations are obtained per selected mobile station.
Then, a base station of the pair of base stations having the minimal OTD value may be selected as the master base station M_BS. Alternatively, one of the base stations of the pair of base stations having an OTD value which is closest to the average value of all obtained OTD values may be selected as the master base station M_BS .
Nevertheless, the most practical way for selecting a master base station M_BS is to choose the base station of an upper cell layer (macro base station) as a master base station for lower layer base stations (micro base stations) . This may be done in case the network is implemented as a microcell and macrocell overlaid system, as is usually the case in densely populated areas (e.g. city areas) where a high traffic capacity of e.g. more than 400 Erl/km2 has to be expected. Such areas are covered by a continuous micro base station network having a density of about 25 micro base stations per square kilometer, for example, with a network of macro cell base stations being overlaid there over (e.g. 5 macro base stations). Thus, when the present invention is implemented, a limited range of geographical area has a
synchronized network (e.g. the range of a macro cell), and the geographical area synchronized could correspond to the area of a railway station, an airport or the like, with the benefits of synchronization being obtained only in this particular area.
Apart from the above, also the implementation of step S205 in Fig. 2 requires attention in implementing the present invention.
That is, the change in the timing of the commanded slave base stations S_BS may reside in changing the frame starting timing (i.e. the starting time of the frame structure) of the slave base stations. Similarly, also the number of frames (occurring in multi-frame transmission) may be changed to establish synchronism (i.e. minimum OTD values) between master and slave base stations. However, prior to change the frame number, the ongoing calls in the concerned cell will have to be handed over beforehand to another base station, since the change of frame number would cause a dropping of ongoing calls. Therefore, a change of frame number should only be chosen when also performing a reset of the concerned base station.
Such a change in timing may be performed immediately. Immediately in this case means abruptly, such that steps S205 to S208 are performed only once to establish synchronism. Such an abrupt change is performed after resetting the system of base stations in the radio network.
Furthermore, in case an asynchronism of the base station develops during network operation and a reset is thus not possible (ongoing calls would otherwise be dropped) , the
adjustment of timing between master and slave base stations can also be performed gradually. In this case, the steps S205 to S208 are repeatedly executed. Such a gradual adjustment uses a long time constant to thereby track the phase, such that ongoing calls are not interrupted and speech quality is affected as little as possible. The time period until synchronization is established depends on the number of times the loop of steps S205 to S208 is executed. When changing the observed timing difference in a gradual manner, the time constant can be influenced by using steps of predetermined size to obtain minimal OTD values.
Furthermore, when the need to change the timing is rather small and/or the time constant is so long that tuning does not cause a significant frequency error, a fine tuning can be performed. The fine tuning resides in changing the frequency of the reference clock of a respective slave base station, with the specified frequency error of the transmitted signal remaining unaffected.
Using the method of the present invention as described above, the synchronization can be easily achieved to be as accurate as less than eight microseconds (8μs). The indicated numerical value for the obtained accuracy is obtained from the capability of a mobile station to handle a maximum variance of delay. In this connection, it was assumed that for a single mobile station MS a propagated radio signal may reach the mobile station after reflection at a reflector (e.g. a building or a mountain) causing 16 μs additional delay or through line of sight. The indicated accuracy is sufficient to ensure a proper utilization of previously discussed enhancement
methods which require synchronized air interfaces Um of a plurality of base stations BS in a radio network.
In addition, with the present invention it is also possible to maintain the same time offset after a software based reset (SW reset) of the base stations, if the base station does not reset its main clock due to the SW reset.
Moreover, although the controller device (base station controller BSC) was described herein above as an individual component of the radio network system , it is evident that the controller device can also be provided inside a base transceiver station BS .
It should be understood that the above description and accompanying figures are only intended to illustrate the present invention by way of example only. The preferred embodiments of the method may thus vary within the scope of the attached claims.