Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
METHOD AND TELECOMMUNICATIONS NETWORK FOR DEACTIVATING OR ACTIVATING A CELL IN SUCH A NETWORK
Document Type and Number:
WIPO Patent Application WO/2011/138346
Kind Code:
A1
Abstract:
The invention relates to a method for deactivation of at least one first cell of a plurality of cells in a telecommunications network. User devices in the at least one first cell are triggered to report measurement information regarding one or more second cells of the plurality of cells to the telecommunications network. The measurement information is received in the telecommunications network (preferably using the still active at least one first cell) and analysed (in the telecommunications network or by an external system), in order to determine whether one or more user devices in the at least one first cell are eligible for being served by a second cell of the one or more second cells when the at least one first cell would be deactivated. When the one or more user devices are determined to be eligible for being served by the second cell of the one or more second cells, the one or more user devices are transferred, i.e. are handed over or perform cell reselection from the first cell to the second cell and the at least one first cell is deactivated.

Inventors:
JORGUSESKI LJUPCO (NL)
LITJENS REMCO (NL)
ZHANG HAIBIN (NL)
Application Number:
PCT/EP2011/057084
Publication Date:
November 10, 2011
Filing Date:
May 04, 2011
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
KONINKL KPN NV (NL)
TNO (NL)
JORGUSESKI LJUPCO (NL)
LITJENS REMCO (NL)
ZHANG HAIBIN (NL)
International Classes:
H04W24/02; H04W24/10; H04W36/22; H04W88/08
Domestic Patent References:
WO2009119699A22009-10-01
WO2008044208A22008-04-17
Foreign References:
EP2141947A12010-01-06
EP2056628A12009-05-06
EP2114093A12009-11-04
Other References:
SUSTAINABLE ENERGY USE IN MOBILE COMMUNICATIONS, June 2007 (2007-06-01)
NEC'S PROPOSALS FOR NEXT-GENERATION RADIO NETWORK MANAGEMENT, February 2009 (2009-02-01)
"Self-configuring and self-optimizing network (SON) use cases and solutions", 3GGP TR 36.902 V9.1.0
Attorney, Agent or Firm:
WUYTS, Koenraad (CH The Hague, NL)
Download PDF:
Claims:
CLAIMS

1. A method for deactivation of at least one first cell of a plurality of cells in a telecommunications network, the plurality of cells defining a coverage area containing a plurality of user devices, the method comprising the steps of:

- triggering user devices in the at least one first cell to report measurement information regarding one or more second cells of the plurality of cells to the telecommunications network;

- receiving and analysing the measurement information regard- ing the one or more second cells to determine whether one or more user devices in the at least one first cell are eligible for being served by a second cell of the one or more second cells when the at least one first cell would be deactivated;

- deactivating the at least one first cell when the one or more user devices are determined to be eligible for being served by the second cell of the one or more second cells.

2. The method according to claim 1, wherein the analysis of the measurement information includes a coverage verification verifying whether a signal level or signal quality at the user device from the second cell meets a signal level condition or signal quality condition. 3. The method according to claim 1 or 2, wherein the analysis of the measurement information includes a quality of service verification verifying whether an estimate of the quality of service obtained in the second cell meets a quality of service condition.

4. The method according to claim 3, wherein the quality of service verification involves the steps of:

- obtaining, from the user devices in the first cell, pilot signal strength and/or pilot quality measurement informa- tion, hereinafter pilot measurement information, related to the one or more second cells;

- assessing the pilot measurement information to assign the pilot measurement information for particular user devices to a particular second cell;

- verifying whether the particular second cell is capable of satisfying a required quality of service for the particular user devices when the first cell is deactivated, the veri¬ fication using the number (Nl) of particular user devices, the number (N2) of user devices already served via the par¬ ticular second cell and the pilot measurement information.

5. The method according to claim 4, further comprising the step verifying whether the particular second cell is capable of satisfying a required quality of service for user de¬ vices already served via the particular cell when the first cell is deactivated and the particular user devices would be handed over to the particular second cell, the verification using the number (Nl) of particular user devices and the number (N2) of user devices already served via the particular second cell.

6. The method according to one or more of the preced¬ ing claims, further comprising the steps of:

- performing an energy consumption reduction analysis or an electromagnetic radiation reduction analysis for the tele¬ communications network to obtain an energy consumption reduction analysis result or an electromagnetic radiation analysis result;

- deactivating the at least one first cell when the energy consumption reduction analysis result or the electromag¬ netic radiation analysis result meets an energy reduction condition or an electromagnetic radiation reduction condition, respectively.

7. The method according to one or more of the preced ing claims, wherein triggering the user devices to report the measurement information regarding the one or more second cells is based on a cell load experienced in the at least one first cell from active user devices.

8. The method according to one or more of the preced- ing claims, wherein the at least one first cell and the second cell of the one or more second cells define a handover area in the coverage area, wherein the trigger to the user devices for reporting measurement information includes a trigger to user devices in the first cell and outside the handover area.

9. The method according to one or more of the preced¬ ing claims, further comprising the step of instructing the user devices in the at least one first cell, preferably using a broadcast channel in the at least first cell, to set or adjust at least one of a measurement threshold to enable the user de¬ vices to obtain measurement information and a reporting

threshold defining a threshold for reporting the measurement in¬ formation . 10. The method according to one or more of the preced¬ ing claims, further comprising the step of requesting and receiving the measurement information from active user devices via signalling channels. 11. The method according to one or more of the preced¬ ing claims, wherein requesting the measurement information regarding the one or more second cells is performed during a limited time interval only, e.g. less than 15 minutes. 12. The method according to one or more of the preced¬ ing claims, wherein the step of triggering user devices in the at least one first cell comprises at least one of the steps of:

- transmitting a reporting identifier in the at least one

first cell;

- manipulating a signal threshold to be applied by the user devices in the at least one first cell; - transmitting a paging message in the first cell instructing idle user devices to obtain and/or report the measurement information;

- temporarily decreasing a location area update period in the at least one first cell.

13. A method in a telecommunications network contain¬ ing a plurality of cells defining a coverage area containing a plurality of user devices, the plurality of cells being in a current state wherein a first cell is an active cell and a sec¬ ond cell is an active cell, the method comprising the steps of:

- retrieving history information about a previous transition from a first previous state wherein the first cell was an active cell and the second cell was an active cell to a second previous state wherein the first cell was an inac¬ tive cell and the second state was an active cell;

- estimating from the retrieved history information whether the first cell can be deactivated. 14. A method in a telecommunications network contain¬ ing a plurality of cells defining a coverage area containing a plurality of user devices, the plurality of cells comprising a first cell as an active cell for a first radio access technology and a second cell is an inactive cell for the first radio access technology and an active cell for a second radio access technol¬ ogy, the method comprising the steps of:

- triggering user devices in the at least one first cell to report measurement information regarding the second radio access technology of the second cell;

- receiving and analysing the measurement information regarding the second radio access technology of the second cell to determine whether one or more user devices in the at least one first cell are eligible for being served by the second cell using the first radio access technology;

- activating the first radio access technology in the second cell when the one or more user devices are determined to be eligible for being served by the second cell using the first radio access technology.

15. A telecommunications network or system configured for performing the method, or steps thereof, according to one or more of the preceding claims.

16. A computer program comprising software code portions configured for, when executed by a processor, performing one or more steps of the method according to one or more of the claims 1-14.

17. A user device configured for operating in a method according to one or more of the claims 1-14.

Description:
Method and telecommunications network for deactivating or acti ¬ vating a cell in such a network

FIELD OF THE INVENTION

The invention relates to the field of telecommunica ¬ tions infrastructures. More specifically, the invention relates to the fields of controlling energy consumption, particularly reducing energy consumption, or electromagnetic radiation, in telecommunications infrastructures comprising wireless access networks .

BACKGROUND OF THE INVENTION

The operation of wireless access networks for enabling wireless communication is highly energy consuming. In view of current environmental concerns, increased attention is paid re ¬ cently to the energy consumption of telecommunications networks.

Various studies have been performed to reduce conven ¬ tional energy consumption in wireless access networks, e.g. by exploring the option of using sustainable energy sources (Erics ¬ son AB White Paper "Sustainable energy use in mobile

communications", June 2007).

With the development of the 3GGP Long Term Evolution (LTE) network, energy saving for the network is also approached in the context of self organizing networks (SONs) . In a White Paper of NEC, dated February 2009, "NEC's proposals for next- generation radio network management", energy is considered as a significant part of the operation expenses of a cellular net ¬ work. It is recognized that the main saving potential resides in using variations in load over time, that allows to switch off parts of the resources, for example during the night. When a complete base station is switched off, other base stations of the access network need to compensate for the reduction in cov ¬ erage area and capacity. This requires coordination between the nodes. A similar use case is described in 3GGP TR 36.902 v9.1.0 "Self-configuring and self-optimizing network (SON) use cases and solutions".

The activation and deactivation of base stations, or cells thereof, or reduction of its operability has implications for user devices (terminals, user equipment (UE) ) in the cover ¬ age area of these base stations or cells.

The current estimation when to switch off/on a base station (or cell) and which base stations to switch off/on is usually based on load and configuration information that might be complemented with handover (HO) statistics. Load measurements can e.g. be performed for one or more cells in the network of a network operator. The network operator also has detailed information on the configuration of e.g. the base station antenna directions and tilting, base stations transmit powers, etc. that can be used, with support of propagation models in order to es ¬ timate the best server areas per cell. Additionally, from network HO counters, the network operator can make HO statistics for the cells. In this way by combining configuration settings, propagation/planning data and HO statistics the network operator can estimate when a particular cell (or base station) can be switched off/on and which remaining cells may provide compensat ¬ ing coverage in the areas of the cells that have been switched off.

The current estimation techniques provide several dis ¬ advantages. The estimations are based on models for the antenna patterns and propagation conditions. These models have intrinsic inaccuracy when compared to the real-life situation due to mod ¬ elling errors and simplifications. Additionally, any change in the antenna configuration, propagation environment, etc. has to be accurately and timely updated in order to maintain some accu ¬ racy of the estimation. This can be a rather demanding task, especially in case of self-optimizing radio access networks that dynamically reconfigure antenna set-up (e.g. tilting or azi ¬ muth), downlink transmission powers, etc.

Furthermore, the planning tool for estimation has to be run in parallel with the changes of the antenna configuration settings, downlink power settings, etc. in order to obtain up- to-date estimation. Running coverage/planning predictions is usually only executed off-line at network roll-out and network extensions. Running such predictions in parallel with the net ¬ work operations and entirely consistent with any change to the network's configuration is a cumbersome task.

Still further, the estimation via planning/propagation tools and configuration data is usually based on some kind of 'average user' assumptions or predictions for the spatial dis ¬ tribution of the user devices and/or the traffic related to the devices. This is another source of intrinsic error due to the uncertainty of the predictions. Even though these spatial traf ¬ fic/device distributions might realistically reflect the average traffic/device spatial distributions, the actual distributions might strongly deviate from the 'average' situation at the mo ¬ ment when a cell has to be switched off or on.

There exists a need in the art for improved control on the effect of the activation/deactivation of base stations, or cells thereof, on the user devices associated with these base stations .

SUMMARY OF THE INVENTION

A method for deactivation of at least one first cell of a plurality of cells in a telecommunications network is dis ¬ closed. The plurality of cells of the telecommunications network define a coverage area containing a plurality of user devices.

User devices in the at least one first cell are trig ¬ gered to report measurement information regarding one or more second cells of the plurality of cells to the telecommunications network. The measurement information is received in the telecommunications network (preferably using the still active at least one first cell) and analysed (in the telecommunications network or by an external system) , in order to determine whether one or more user devices in the at least one first cell are eligible for being served by a second cell of the one or more second cells when the at least one first cell would be deactivated.

When the one or more user devices are determined to be eligible for being served by the second cell of the one or more second cells, the one or more user devices are transferred, i.e. are handed over or perform cell reselection, from the first cell to the second cell and the at least one first cell is deacti ¬ vated .

It should be noted that in an embodiment of the inven ¬ tion, handover is performed prior to deactivating the at least one first cell. Cell reselection may take place either before, at or after deactivating the at least one first cell.

A telecommunications network wherein this method can be performed is also disclosed.

A computer program containing software code portions, possibly run on different systems, for performing the method is also disclosed.

A user device, particular a user device in an idle state, configured for participating in the method and telecommu ¬ nications network is also disclosed.

By instructing a substantial fraction (possibly all) user devices in the first cell to report measurement information regarding one or more second cells observed by the user devices and analysing the received measurement information, an almost real-time assessment of opportunities to deactivate a particular cell or base station is obtained. By deactivating the particular cell or base station, energy consumed by and/or electromagnetic radiation radiated by the telecommunications network may be re ¬ duced .

It should be noted that in the present disclosure a cell is considered x active' when the cell is configured to pro ¬ vide to a user device substantially all services it usually provides during normal operation. For example a user device can connect to the base station responsible for defining the cell and the base station supports the traffic flow for voice and/or data services. A cell is considered x inactive' or λ deactivated' when the cell is configured to provide to a user device no ser ¬ vice or to provide only a limited set of services compared to what it usually provides during normal operation. For example the base station does not support new connection requests, the base station does not support a traffic flow for voice and/or data services, the base station supports only a limited set of radio resources or of mobility management messages and/or a user device cannot connect to that base station. Note that an inac ¬ tive cell is not necessarily free from any signal from the base station. Signals such as broadcast signals or pilot signals may still be present and some information may be exchanged between such a cell and a user device. Of course, a cell is considered inactive when no signals are transmitted in this cell from the base station previously responsible for this cell.

Further embodiments of the invention are defined in the dependent claims.

An alternative method for analysing whether or not to deactivate a cell includes a method in a telecommunications net ¬ work containing a plurality of cells defining a coverage area containing a plurality of user devices. The plurality of cells are in a current state wherein a first cell is an active cell and a second cell is an active cell.

History information is retrieved about a previous tran ¬ sition from a first previous state wherein the first cell was an active cell and the second cell was an active cell to a second previous state wherein the first cell was an inactive cell and the second state was an active cell. As an example, if it is considered to deactivate the first cell at a particular time of day/week, information is retrieved about the effect of switching off the first cell in the past at that particular moment. The information could e.g. relate to the number of handovers from the first active cell to the second cell. It may then be esti ¬ mated from the retrieved history information whether the first cell may be deactivated. The first cell is deactivated based on the estimation that user devices may be transferred to the ac ¬ tive second cell.

Another aspect of the present disclosure includes a method for activating a cell in a telecommunications network

A method in a telecommunications network containing a plurality of cells defining a coverage area containing a plural ¬ ity of user devices is disclosed. The plurality of cells

comprise a first cell as an active cell for a first radio access technology and a second cell as an inactive cell for the first radio access technology and an active cell for a second radio access technology. Examples of radio access technologies, also abbreviated as RATs, include GSM, UMTS and LTE .

User devices in the at least one first cell are trig ¬ gered to report measurement information regarding the second radio access technology of the second cell. The measurement in ¬ formation regarding the second radio access technology of the second cell is received in the telecommunications network (e.g. via the first cell using the first RAT or via the second cell using the second RAT) to determine whether one or more user de ¬ vices in the at least one first cell are eligible for being served by the second cell using the first radio access technol ¬ ogy.

The first radio access technology is activated in the second cell when the one or more user devices are determined to be eligible for being served by the second cell using the first radio access technology.

In order to improve the accuracy of estimations for the first RAT in the second cell based on measurement information for the second RAT in the second cell, the second cell for the first RAT and second RAT are co-sited.

It should be noted that, separate from or in addition to considerations of energy conservation, similar considerations apply to the reduction of electromagnetic radiation produced by a base station. For example, for health reasons it is advisable to limit the amount of electromagnetic radiation to a level that is strictly necessary for the operation of the base station.

When the level of electromagnetic radiation can be reduced or when a base station can be switched off, this may be well- appreciated by the population in the vicinity of the base sta ¬ tion. In addition to health concerns, there may also be

technical reasons to restrict the electromagnetic radiation, and therewith the possibility and level of interference experienced in adjacent areas.

Hereinafter, embodiments of the invention will be described in further detail. It should be appreciated, however, that these embodiments may not be construed as limiting the scope of protection for the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS

In the drawings:

FIGS. 1A and IB are schematic illustrations of base stations providing active cells and inactive cells;

FIGS. 2A and 2B are schematic illustrations of a first and a second base station defining a cell containing a plurality of user devices to be transferred to neighbouring cells of an ¬ other base station;

FIG. 3 is a schematic illustration of signal level and signal level threshold depicting a mechanism for controlling from the network whether user devices perform measurements according to an embodiment of the invention;

FIG. 4 is a flow chart for deciding deactivation of a cell of a telecommunications network according to an embodiment of the invention;

FIGS. 5-7 are flow charts illustrating steps of a method for participation of idle user devices;

FIG. 8 is a schematic illustration of a wireless user device configured for being operable in the system of FIGS. 2A and 2B;

FIG. 9 is a flow chart for QoS verification for user devices to determine eligibility for transfer to another cell;

FIG. 10 is a flow chart for deciding deactivation of a cell of the telecommunications network according to an embodi ¬ ment of the invention;

FIG. 11 is a flow chart for deciding activation of a cell of the telecommunications network according to another em ¬ bodiment of the invention; and

FIG. 12 is a schematic illustration of a multi-RAT net ¬ work wherein a cell is activated.

DETAILED DESCRIPTION OF THE DRAWINGS

FIG. 1A is a schematic illustration of base stations BS A - BS G providing a plurality of active cells. Each base sta- tion location comprises three sectors, thus covering an area around the base station location. For example, cell Al of base station location BS A serves its best server area marked x al' . Similarly, cells A2 and A3 (not shown) of the same base station location BS A serve their best server area marked x a2' and x a3' respectively. A same set-up is shown for the other base station locations BS B to BS G. In FIG. 1A, the regular pattern of base stations and sectors leads to the commonly known hexagonal of coverage areas .

At some point in time, cell Al of base station BS A may experience a low load of active user devices. In such a situa ¬ tion, it may be profitable for energy consumption reduction or other reasons to deactivate, e.g. to completely switch off, cell Al or even all cells A1-A3 at this base station (illustrated by the white server area al in FIG. IB) . Direct deactivation of one or more cells at the base station BS A may affect user devices served by this cell. The effect of switching off one or more cells on these user devices should be taken into account. In particular, it should be analysed if it is feasible to transfer the current (and in the near future anticipated) traffic from the cell of base station BS A to neighbour cells B2 and C3 and, if so, to which neighbour. Preferably, it should be verified that for (substantially) all user devices currently served by the cell Al of base station location BS A there is indeed an al ¬ ternative. Furthermore, it may be verified whether a reasonable quality of service (QoS) can still be obtained for the user de ¬ vices taken over by other cells and, similarly, whether a reasonable QoS can still be maintained for the user devices al ¬ ready being served by one of the other cells. The measure of deactivating cell Al to reduce energy consumption is weighed against the implications for the user devices for which there is no alternative and furthermore for the QoS likely to be experi ¬ enced by the user devices involved. In other words, a base station deactivation algorithm is used for analysing whether traffic can be transferred from the cell Al of base station BS A to cells B2 and C3 of base stations BS B and BS C, respectively, and whether acceptable QoS levels can be maintained to decide whether or not to deactivate cell Al of base station A.

Referring now to FIGS. 2A and 2B, a situation is depicted wherein a cell of base station A is considered as a candidate to be switched off, for example, because at that mo ¬ ment the number of active user devices served by the cell in base station A is rather low. For example, as presented in FIG. 2A, there are two active user devices 1 and 2 (i.e. devices that have a signalling/traffic connection with the base station A) .

The number of idle user devices 3 to 9 (i.e. user de ¬ vices that do not have a signalling/traffic connection with the base station A but only camp on this base station A) is much larger .

Apart from the distinction between idle and active user devices, a further distinction between the user devices 1-9 can be observed.

Active user device 2 is in a handover region and performs signal measurements to prepare for handover to the cell of base station B. In this situation, user device 2 reports meas ¬ urements about the cell of base station B and possibly about other cells to the telecommunications network if certain condi ¬ tions are met (e.g. if the level of the pilot signal received from base station A is below a threshold) . Different from active user device 2, active user device 1 is not in a handover region and does not measure and report signals from other cells (e.g. because the pilot signal level received from base station A is above a threshold) . It should be noted that measurements made by a user device about base station B may involve the same radio access technology or a different radio access technology than the radio access technology (RAT) of base station A.

A further distinction is also applicable to the idle user devices 3-9 in FIG. 2A. The idle user devices may also measure neighbour cells, such as the cell (s) of base station B. Idle user devices 5, 6 and 8 verify whether they should camp on a cell of base station B instead of the cell of base station A. Idle devices 3, 4, 7 and 9 do not perform such a verification. Again, the same or different RATs may apply for base stations A and B .

In order to assess whether the cell of base station A can be deactivated, it may be advantageous to instruct all, or a substantive number of, user devices 1-9 to report measurement information about neighbour cells. In particular, active user devices not finding themselves in the handover region and idle user devices should be instructed to measure and report informa ¬ tion received about neighbour cells (e.g. pilot level strength, estimated path loss, etc. together with the cell ID) as these user devices do not measure and report on their own account and, in addition, are more likely to be affected severely if base station A is deactivated.

Referring now particularly to FIGS. 2B and 4, an embodiment of the invention operates as follows.

As a first step, base station A triggers active and idle user devices 1-9 in cell A to perform measurements on one or more neighbouring cells. Various methods to trigger the user devices will be discussed below.

In a second step, the user devices 1-9 report the meas ¬ urement information, e.g. to base station A.

The measurement information (possibly aggregated per user device) can be represented as a table as shown in FIG. 2B. An estimate of the contribution of the cell to the coverage of the user device by this cell and QoS achievable for the user de ¬ vice can be made. Neighbouring cells can be ranked and a

selection is made for the highest ranked neighbour cells.

After collecting the neighbour pilot signal level meas ¬ urements from all (active and idle) user devices 1-9, it may be determined or estimated how many user devices do not have any (intra- or inter-RAT) neighbour with a pilot signal level that satisfies a (RAT-specific) minimum threshold (e.g. for cover ¬ age) . The coverage verification may e.g. be considered passed if the number or the fraction of the user devices currently in the cell of base station A which indicate that not at least one suf ¬ ficient alternative pilot signal level is found, is lower than an operator-defined threshold (e.g. less than three devices or lower than 1%, etc.) .

Similarly, after collecting the estimated path loss measurements (or after determining values of path loss from the collected pilot level measurements and related neighbour cell information) , preferably for those neighbours that satisfy the minimum threshold on measured pilot signal level, from all ac ¬ tive user devices, it may be determined or estimated which QoS is achievable for each of the active user devices 1, 2 when these would be served by a particular neighbour. For this purpose the telecommunications network can also use load

information signalled from the (intra- or inter-RAT) neighbours of interest and the current resource usage/availability in those neighbours. The QoS verification may e.g. be considered passed when the estimated resulting QoS level satisfies an operator- defined minimum QoS level.

As a next and final step, when the processing is final ¬ ized, a decision is taken whether to switch off one or more cells of base station A. The decision may also take account of an energy consumption reduction analysis result obtained from running an energy consumption reduction analysis algorithm. If the result meets an energy condition, indicating significant energy savings from deactivating one or more cells of base station A, the one or more cells of base station A can actually be deac ¬ tivated. The decision may also be indicated towards neighbour base station B. Base station B may possibly adapt its configura ¬ tion prior to deactivating a cell of base station A in order to prepare for the expected additional traffic and/or extension of the area to serve.

Instead of or in addition to energy conservation considerations, also the possibilities to reduce electromagnetic radiation could be taken into account. In particular of electro ¬ magnetic radiation reduction, different weights could be

associated with different areas, for example depending on (ex ¬ pected) population density. As mentioned above, several methods have been envisaged to instruct user devices to perform measurements and/or to re ¬ port the measurement information.

As schematically illustrated in FIG. 3, a threshold pi ¬ lot signal level T_meas_neighbour may apply within the cells of base station A. The value of this threshold is generally broad ¬ cast in a cell as system information. If a user device 1-9 in cell A measures cell A' s pilot signal level to be below that threshold, the user device starts measuring neighbour cells, e.g. one or more cells of base station B. In order to instruct all (active or idle) user devices in cell A to measure neighbour cells, the network can adjust, temporarily, the corresponding threshold T_meas_neighbour in cell A to a rather high signal strength value, so that substantially all user devices in cell A (even the user devices that are not preparing for handover or are not considering cell reselection, i.e. user devices outside the handover region and close to base station A) will measure neighbour signal strengths.

In principle, the threshold may be set to different values for idle user devices and for active user devices. Simi ¬ larly, different values may also be used for measuring intra-RAT or inter-RAT neighbours. For reasons of clarity, FIG. 3 only il ¬ lustrates a single threshold T_meas_neighbor .

After having performed measurements on neighbour cells, the user devices should report the measurement information to the telecommunications network, e.g. to base station A.

Active user device 2, which is in the handover region and is preparing for handover, reports neighbour signal

strengths already (controlled and formatted by the network) in order for the network to initiate the handover towards the de ¬ sired target cell. Active user device 1, which is not in the handover region and does not prepare for handover (but may have been triggered to perform neighbour measurements, e.g. by ex ¬ plicit signalling or by temporarily adjusting the threshold T_meas_neighbour as described above) can be explicitly signalled (as there is a signalling connection between active user device 1 and the network via cell A) to report the neighbour measurements. Alternatively, the network can adjust a reporting threshold for the active user device (using e.g. the system in ¬ formation provided via the broadcast channel) such that more active user devices will report neighbour measurements.

Whereas the manipulation of the threshold pilot signal level T_meas_neighbour is applicable for all user devices 1 in the cell, it should be appreciated that individual active user devices can also be instructed to perform measurements via the existing connection with the network. Of course, reporting of the measurement information may take place via the existing con ¬ nection as well.

It should further be noted that for some RATs, such as UMTS, active user devices 1 always collect measurement informa ¬ tion on neighbours on the same frequency. In such cases, the reporting conditions for delivering the measurement information to the network may be manipulated.

Next, exemplary methods will be discussed for instruct ¬ ing idle user devices 3-9 to perform and report measurements on neighbour cells and to report this information to the telecommu ¬ nications network.

In a first method, illustrated in the flow chart of FIG. 5, a reporting identifier, also referred to as reporting flag, is broadcast in the cell Al of FIG. 1A.

In step 5-1, the system information (SIB) to be broadcast on the broadcast channel of cell Al comprises a reporting flag. In cell Al, a paging message may be issued indicating modified system information. This paging message is also re ¬ ceived by idle user devices camping on the active cell Al .

An idle user device T, detecting the reporting flag, is configured (see FIG. 8) to report information about its identity (e.g. IMSI) and about n best cells observed by the user device, possibly including signal strength, signal quality, path loss estimate, etc. for these best cells. In doing so, the user de ¬ vice in detecting the reporting flag may adjust temporarily a signal level/quality measurement threshold to enable the meas ¬ urements. The report is sent once on first detection of the reporting flag being set. In step 5-II, the reports of idle user devices are re ¬ ceived in the telecommunications network via cell Al and

processed in a processing system of the telecommunications net ¬ work. The reporting flag set in cell Al is a signal for the idle user device T to assemble and transmit the report via same cell Al .

After sufficient information has been obtained in step 5-III in the cell Al, the system information (SIBs) may be reverted to normal operating conditions (which also includes removal of the reporting flag) . A paging message may again be issued indicating modified system information.

In a second method, illustrated in the flow chart of FIG. 6 as step 6-1, a measurement-and-report paging message, hereinafter denoted as paging message, is transmitted in the cell Al instructing the idle user devices T to obtain and report information about the neighbour cells. The user device is configured, in response to receiving the paging message of step 6- I, to report information about the identity of the user device

(e.g. IMSI) and about n best cells observed by the user device, possibly including signal strength, signal quality, oath loss estimate etc. for these best cells. The report is sent once af ¬ ter receipt of the paging message.

The paging message may optionally comprise configura ¬ tion parameters for the idle user device T. Examples are

parameters that further specify the measurements to be made

(which terminals should measure, which cell (s) to be measured, which RAT to be measured, which characteristics, etc.) and pa ¬ rameters that further specify the report to be provided (how many of the best cells to be reported, which quantities - e.g. signal level/quality, path loss estimates - to be reported, des ¬ tination of the report, etc.) . The paging parameters may

alternatively or in addition, include a reference where the

(further) parameters may be obtained, e.g. a reference (link) to a broadcast channel or the like, for example when the set of pa ¬ rameters is impractically large to be sent in the paging message itself . A user device T receiving configuration parameter values in (or associated with) the paging message applies the received values when performing the measurements and the report ¬ ing as commanded by the paging message. For parameters not included in the paging message, default values may apply, e.g. preconfigured in the user device or received as system informa ¬ tion. It may be that, in case of a conflict with an existing parameter having been assigned a value, the value received in the paging message is only applied for the measurements and re ¬ port (s) related to the present method. Other procedures continue to apply the already assigned value and are not affected by the value in the paging message.

In step 6-II, the reports of idle user devices T are received in the telecommunications network via cell Al (by de ¬ fault, as being the originator of the paging message) (or via different cell (s) when the paging message instructed the user devices to do so) and processed in a processing system of the telecommunications network.

In a third method, illustrated in the flow chart of FIG. 7, a location area (LA) update period (or a similar period) is decreased in at least cell Al . The embodiment is advantageous as it requires no modifications to currently existing networks.

In the cell Al, in the system information block (SIB) to be broadcast on the broadcast channel of the cell Al, a rela ¬ tively short value (e.g. 30 s, 1 min) is set for the periodic location area update (instead of a nominal value, e.g. 30 min, 1 hour, infinite) . This is illustrated in step 7-1. A paging mes ¬ sage may be issued in the cell Al indicating modified system information .

An idle user device T camping on the cell Al reads the modified SIB and starts providing frequent (e.g. every 30 s, 1 min) periodic LA updates. An LA Update message also includes in ¬ formation about the identity of the user device T.

The telecommunications infrastructure (more particu ¬ larly a processing system thereof) receives the LA Update messages and may derive (e.g. every 30 s, 1 min) how many and which user devices T are currently camping on the cell Al . This provides information about the initial situation regarding idle user devices in the cell. When this information is not required, this analysis and the preceding steps may be omitted.

In step 7-II, in the cell Al, in the system information to be broadcast on the broadcast channel of the cell, the pa ¬ rameters determining the search behaviour of idle user devices T and/or parameters affecting the ranking of the cell Al are modi ¬ fied. Most modern wireless access networks such as GSM, UMTS, LTE, provide various parameters to affect the process of cell evaluation for cell selection and reselection performed by idle terminals. The (preferably idle-mode specific) parameter (s) af ¬ fecting the ranking of cell Al is set to a value such that cell Al is, as much as possible, ranked away from the top of the list indicating the x best cell' . In this manner, it is unlikely that idle terminal T, currently camping on cell Al, will still rank cell Al as best cell in a next cell reselection procedure. Also, if the previous (optional) steps 7-1 were omitted, a rela ¬ tively short value (e.g. 30 s, 1 min) is set for the periodic location area update (instead of a nominal value, e.g. 30 min, 1 hour, infinite) . In the cell, a paging message may be issued in ¬ dicating modified system information. It should be appreciated that for some RATs (e.g. UMTS) an idle user device T will exe ¬ cute the cell reselection evaluation process when triggered by a modification of the information on the BCCH used for the cell reselection evaluation procedure (i.e. when this particular SIB is modified) .

As a consequence, an idle user device T for which the search criterion applies, measures its neighbour cells. For some RATs (e.g. UMTS) a user device will perform such measurements regardless the setting of the search criterion.

An idle user device T for which the specified condi ¬ tions are satisfied performs cell reselection. The condition for cell reselection may be based on ranking the cell Al and all neighbour cells. By adjusting the parameter (s) affecting the ranking of cell Al, it is envisaged that cell Al is not (no longer) the highest-ranked cell and that the idle user device T will make a cell reselection to the highest-ranked cell which is different from cell Al .

An idle user device T persisting to camping on the cell Al continues providing frequent LA Updates messages to the cell Al . An idle user device T which has reselected a neighbour cell will adhere to the LA Update regime as broadcast by that cell (which may be a nominal value, e.g. 30 min, 1 hour, etc.) . Op ¬ tionally, for example concurrent with step 7-1, neighbour cells of cell Al are also configured for frequent LA Updates as indi ¬ cated above and in FIG. 7. In that case, an idle terminal Ti which has reselected a neighbour cell Bl will provide its fre ¬ quent LA Update messages to the relevant cell Bl resulting in more prompt information to the network about the reselection of terminals towards that neighbour cell.

After a transitional period (whose length may depend on the duration of the cell reselection procedure and the LA Update period) , the processing system of the telecommunications infra ¬ structure may at least derive in step 7-III, from the received LA Update messages still being received from cell Al, how many and which idle user devices remain camped on the cell Al, indi ¬ cating that these user devices were unable to reselect an alternative cell. By, in addition, using the option of step 7-1 and/or the option of frequent LA Updates for cell Al's neighbour cells, the processing system may also derive, from the LA Update messages received previously via cell Al and from the LA Update messages received currently via any of its neighbour cells, how many and which idle user devices have reselected a neighbour cell (and which cell) and how many and, possibly, which idle user devices have reselected from the cell Al to other cells. This information is indicative of the situation regarding idle user devices T when cell Al would be deactivated. Also if the options outlined above (the option of step 7-1 to obtain infor ¬ mation about the initial situation in cell Al and the option to obtain information about to which cells the idle terminals rese ¬ lected) are not used, the analysis provides at least an estimate about the number (and identity) of idle user devices T to be ad ¬ versely affected when cell Al would be deactivated. The information obtained in step 7-III may be taken into account in the evaluation by the processing system about whether or not deactivate cell Al . The processing result can be analyzed against the at least one deactivation condition

After sufficient information has been obtained in step 7-III, the system information in the neighbour cells may be reverted to normal operating conditions, as far as applicable (step 7-IV) , including reverting the LA update periods to nominal values. A paging message may be issued indicating modified system information. When cell Al is decided not to be deacti ¬ vated, the system information in the cell Al may be reverted to normal operating conditions (step 7-IV) . A paging message may be issued indicating modified system information.

In the above methods, when it is decided to deactivate a given active cell, the idle user devices T camping on this cell are preferably informed beforehand, as to allow them to perform a cell reselection and camp on another active cell. Several solutions exist for this. The more elegant solution is (assuming that active user devices are handed over to suitable neighbour cell (s) , e.g. using existing and conventional methods) to indicate the cell as not being suitable to camp on. Most mod ¬ ern radio access technologies have facilities, for example, to set access class limitations, to indicate that the cell is not intended for user traffic, etc. For example, set the x cell barred' indicator in the system information of cell Al . Depending on the applicable access technology, a paging message may be issued in the cell Al indicating modified system information. Once the idle user devices T have detected that the cell they currently camp on is barred, they will perform a cell reselec ¬ tion. Another, less elegant, way is to simply switch off the cell and let the idle user devices discover (typically within no more than a few seconds) that the cell is no longer active, upon which they will perform a cell reselection.

Whereas the disclosed method and telecommunications network may use existing measurement and reporting capabilities of wireless user devices, the capabilities of wireless user de ¬ vices should be adapted to participate in these embodiments. Examples have been indicated above and include the recognition, processing and actions performed for cell reselection, in response to the reporting identifier (reporting flag) and/or the measurement-and-reporting paging message of the above-described methods. Most of these enhanced capabilities will be obtained by software modifications in the wireless user devices. Therefore, FIG. 8 is a highly schematic illustration of a wireless user de ¬ vice T comprising storage S and processor uP configured for storing and operating the modified software.

It should be noted that neighbour measurements (pilot signal strength/quality, path loss estimate) and reporting for active user device 1 that is not preparing for handover and for all idle user devices puts an extra burden on the user devices and increases its battery usage. Therefore, this triggering by the network preferably is performed during a limited time inter ¬ val, e.g. during a few minutes before a decision to switch off base station A in order to verify that the user devices in the cell (s) of base station A are eligible for being served by in ¬ tra- or inter-RAT neighbour cells. Once sufficient measurements have been made, the nominal threshold setting should be used and there is no longer a measurement and reporting overhead for the active and idle user devices. This applies in particular to the third method for the idle user devices, whereas the overhead in ¬ duced by the first or second method is predominantly event- related .

As indicated in FIG. 4, a QoS verification may be part of the step of analysing the measurement information received from the active user devices. In particular, it may be verified whether an appropriate (e.g. a minimum) QoS can still be ob ¬ tained for one or more user devices once a cell is deactivated and the user devices are handed over to the most suitable second cell. Generally, it should be appreciated that QoS verification may be direction dependent, i.e. the QoS requirements may be different for uplink (UL) en downlink (DL) and QoS verification is preferably performed separately for each direction.

FIG. 9 shows a general flow chart for QoS verification in a telecommunications network for a plurality of active user devices in a cell that is a candidate to be deactivated. A cell may be considered a candidate to be deactivated since the cover ¬ age check (see FIG. 4) for user devices in this cell was

successful, i.e. most user devices proved capable of being cov ¬ ered by neighbouring second cells.

In step 9-1, active user devices are triggered to re ¬ port measurement information, particularly the strength or quality of a pilot or reference signal (pilot measurement infor ¬ mation) , to the telecommunications network. Of course, the pilot or reference signal used can be same signal used for the cover ¬ age check as explained above with reference to FIGS. 3 and 4.

In a next step, step 9-II, the measurement information is analysed in order to obtain knowledge of the preferred (best) candidate cell for substantially each user device in the cell that is considered for deactivation. The pilot measurement in ¬ formation is assigned to this candidate cell. This step also results in information on the number of user devices that will potentially be served by each of the candidate second cells in addition to the known number of user devices that are already served by these cells.

It should be noted that the selection of the user de ¬ vices regarding the best potentially serving cell may e.g. be performed in the base station of the first cell or in another processing device within or external to the telecommunications network. The measurement results for the selected user devices may e.g. be forwarded to the base station responsible for a can ¬ didate second cell. Alternatively, the measurement results for all user devices may be transferred to all base stations respon ¬ sible for the second cells, such that these base stations determine which user devices can potentially be served by which one of their cells.

Then, in a final step 9-III, it is assessed whether an appropriate QoS can be achieved for the potentially additional user devices to be served by the cell, using the measurement in ¬ formation (to derive e.g. the resource requirements) and the total number of user devices in the cell, assuming that the po ¬ tentially additional user devices are indeed served by the cell. Below, the general approach for QoS verification will be described in further detail for two specific scenarios. It should be appreciated that the general QoS verification approach is applicable to other scenarios as well. Again, the first cell is assumed to be the candidate cell for deactivation and the one or more second cells are cells to which the active user devices may be handed over.

The first scenario involves a QoS verification for a HSDPA or LTE network in the downlink (DL) direction. For HSDPA networks, the pilot signal for the one or more second cells is given by the Common Pilot Channel (CPICH) ; for LTE, the reference signal (RS) is used.

For step 9-1, active user devices are triggered to re ¬ port the pilot quality as measured for the one or more of the second cells to the first cell. For HSDPA, the pilot quality can be expressed as the energy per chip over interference plus noise density (Ec/NO) , whereas for LTE, the pilot quality can be ex ¬ pressed as the reference signal received quality (RSRQ) .

For step 9-II, the active user devices are selected as eligible for being served by a particular second cell on the ba ¬ sis of the reported pilot quality. As an example, active user device 1 in FIG. 2A is selected as a user device that is eligi ¬ ble for being served by a particular cell of base station B. For each potentially serving cell, the number Nl of potentially ad ¬ ditional user devices 1 is thus known.

Then, for step 9-III it is assessed whether in the one or more second cells an appropriate QoS can be achieved for the potentially additional user devices to be served by the cell, whereas QoS requirements can also be satisfied for the currently served user devices once the additional user devices would in ¬ deed be handed over.

HSDPA and LTE networks have shared transport channels. In systems with shared transport channels, the experienced aver ¬ age throughput of a given user device can be estimated by the average bit rate as experienced during time frames when the user device is actually scheduled for transmission, multiplied by the fraction of time the user device is indeed scheduled. Assuming some form of fair resource sharing, as typically applied, this time fraction is equal to 1/N, with N the number of active user devices .

The following information elements are available for a best second cell. It is assumed that the second cell already serves N2 user devices. Then, an experienced bit rate for these user devices when scheduled for transmission on the shared transport channel is known.

Furthermore, a mapping or table is assumed to be avail ¬ able relating a particular pilot quality to an achievable bit rate when scheduled for transmission. Such a mapping or table can be updated and adjusted based on based on live network meas ¬ urements .

Moreover, for each user device session, the (minimum) downlink QoS requirement is known, either from the session- specific QoS profile (e.g. based on HLR (Home Location Register) information) or as an operator policy or target parameter conveyed e.g. via the OMC (Operations and Maintenance Center) . The downlink QoS requirement may also be known from the PDP context. The QoS requiremement may be zero, i.e. a minimum QoS require ¬ ment does not exist. The QoS requirement may be a minimum or threshold throughput R T H, e.g. expressed in bit per second

(bit/s), for a user device for a particular service.

From these information elements, QoS verification can be performed for both the Nl potentially additional user devices intended to be handed over to the second cell and for the N2 user devices already served by the second cell.

For the potentially additional user devices of the sec ¬ ond cell, the downlink throughput for the potentially newly added user devices is estimated by dividing the estimated bit rate when scheduled for transmission, by the updated number N1+N2 of served user devices in the considered second cell.

Herein, the estimated bit rate when scheduled for transmission is based on the reported pilot quality in combination with the aforementioned mapping or table. The QoS verification entails a comparison of these throughput estimates with the QoS require ¬ ments, such as the minimum required throughput RTH - For the already serviced user devices, the estimated (reduced) downlink throughput experienced by the currently served user devices in the second cell can be determined by di ¬ viding their experienced bit rates when scheduled for

transmission, by the updated number of served user devices

N1+N2. The QoS check entails a comparison of these throughput estimates with the QoS requirements, such as the minimum re ¬ quired throughput R T H -

It should be appreciated that the QoS verification may need to be performed several times, considering different sub ¬ sets of potentially handed over (candidate) user devices. This may be needed if a QoS verification points out that not all can ¬ didate user devices can be served with adequate QoS by the considered second cell. In such case, (at least) one such user device may be rejected and, since fewer user devices are then estimated to share the second cell's resources, the QoS check for the remaining user device may need to be performed again.

Eventually, the above method results in one or more subsets of candidate user devices that can indeed be accommo ¬ dated by second cells. Combining this information for all second cells then provides sufficient QoS-related input for deciding whether or not the first cell considered for deactivation may indeed be deactivated.

It is noted that in the above method, the pilot quality measurements reported by the candidate user devices may underes ¬ timate the pilot quality actually experienced after deactivating the first cell, since the measurements on the second cells are are performed while the first cell is still active and the meas ¬ urements, hence, include some interference experienced from that first cell. If the first (currently serving) cell is indeed de ¬ activated (as targeted) , this interference no longer exists and the true pilot quality (for each candidate second cell) will be somewhat higher. Effectively, the above method thus provides a somewhat pessimistic result on the attainable QoS. This implic ¬ itly establishes a margin for estimation inaccuracies.

Experience should point out whether this margin is adequate or perhaps too conservative or speculative, in which case a more explicit (positive or negative) margin may need to be applied.

A second scenario relates to QoS verification in an LTE network in the uplink (UL) direction.

For step 9-1, active user devices are triggered to re ¬ port the reference signal received power (RSRP) as measured for one or more for the one or more of the second cells to the first cell. It is noted that the RSRP may also be used for performing the coverage check.

For step 9-II, the active user devices are selected as eligible for being served by a particular second cell. The first cell may then forward the measurements to the respective second cells towards which the active user devices may be handed over. Knowing the reference signal transmit powers for the second cells and assuming channel reciprocity (i.e. it is assumed that the path loss in the UL and the DL direction is identical) , the RSRP measurements and reference signal transmit power are used to estimate the uplink path loss from the given user device to the considered second cell. In other words, from the RS transmit power at the base station and the RSRP measurement by the user device, the downlink path loss is estimated. The same value is used as an estimate for the uplink path loss.

Then, for step 9-III it is assessed whether in the one or more second cells an appropriate QoS can be achieved for the potentially additional user devices to be served by the cell, whereas QoS requirements can also be satisfied for the currently served user devices once the additional user devices would in ¬ deed be handed over.

The assessment can be based on the following informa ¬ tion elements.

First, the number of currently available physical re ¬ source blocks (PRBs) is assessed, as well as the interference level that is experienced in each PRB and the current value of the target received power density Po per PRB.

A further information element is a mapping of an estimated signal-to-interference ratio (SINR) to an appropriate modulation and coding scheme (MCS) , and hence the corresponding data rate, is known from e.g. live network experience, labora ¬ tory experiments or simulations.

Moreover, for each user device session, the (minimum) uplink QoS requirement is known, either from the session- specific QoS profile (e.g. based on HLR (Home Location Register) information) or as an operator policy or target parameter conveyed e.g. via the OMC (Operations and Maintenance Center) . The uplink QoS requirement may also be known from the PDP context. The QoS requiremement may be zero, i.e. a minimum QoS require ¬ ment does not exist. The QoS requirement may be translated to a required number of PRBs .

For the user devices currently active in the considered second cell, the number of PRBs can be determined based on the currently experienced throughput, the current number of assigned PRBs, the experienced SINR levels per assigned PRB and the throughput (QoS) requirement. With this information, the minimum set of required PRBs to satisfy the QoS requirement of the user device can be determined. Having derived the required number (or set) of PRBs per active user device, it is immediately known how many and which uplink PRBs are available for assignment to po ¬ tentially additional user devices from the first cell which is considered to be deactivated.

Then for each such potentially additional user device, the number/set of required PRBs is estimated that is needed to satisfy its QoS requirement. This can be done by considering different (increasing) sets of assigned PRBs, estimate the ef ¬ fective aggregate SINR level (based on Po, the measured

interference levels per PRB, and e.g. an exponential effective SINR mapping (EESM) method for aggregation of the estimated PRB- specific SINRs) , choose the correspondingly estimated most suit ¬ able MCS from an SINR-to-MCS mapping (which can be determined or updated based on network experience) , and determine the corre ¬ sponding aggregate bit rate (which follows directly from the MCS) . With this approach, the minimal set of PRBs needed to achieve an aggregate bit rate that exceeds the different user device specific uplink QoS requirements can be derived for each user device. Per user device, it is then verified whether the target received power level Po per PRB can actually be achieved, given the maximum transmit power of the user device and the estimated path loss (calculated by subtracting the RSRP from the transmit power of the Reference Signal) towards the considered second cell (coverage check) . In other words, it is checked whether the transmit power of the user device is sufficient to achieve the targeted Po level for all (potentially) assigned PRBs to satisfy the QoS requirement. Basically, the required transmit power, i.e. Po * the required number of PRBs * estimated path loss (all in linear units) , should not exceed the maximum transmit power P max of the user device.

Knowing the required number of PRBs for all candidate user device, and having performed the coverage check, it may be verified whether all or only a subset of candidate user devices can be handed over to the considered second cell, given the available set of PRBs. Eventually, the method may result in one or more subsets of candidate user devices that can indeed be adopted by each neighbour cell. Combining this information for all second cells then provides sufficient QoS-related input for deciding whether or not the first cell considered for deactiva ¬ tion may indeed be deactivated.

FIG. 10 shows an alternative method for deciding whether or not a particular cell A can be deactivated using history information. History information is retrieved about a previous transition from a first previous state wherein the first cell was an active cell and the second cell was an active cell to a second previous state wherein the first cell was an inactive cell and the second cell was an active cell. As an ex ¬ ample, if it is considered to deactivate the first cell at a particular time of day/week, information is retrieved about the effect of switching off the first cell in the past at that par ¬ ticular moment. The information could e.g. relate to the number of handovers from the first active cell to the second cell. It may then be estimated from the retrieved history information whether the first cell may be deactivated. The first cell is ac- tivated based on the estimation that user devices may be trans ¬ ferred to the active second cell.

Consider now the situation of FIG. IB, wherein base stations BS B to BS G are active base stations and BS A is an inactive base station for at least one RAT. Increased network load for base stations BS B and/or BS C may ultimately result in congestion in the cells B2, C3 of these base stations, such that activation of cell Al for the specified RAT is desirable. Since base station BS A is inactive for a particular RAT, a problem arises as to how much traffic can be shifted to cell Al .

The method depicted in the flow chart of FIG. 11 in ¬ volves using inter-RAT measurements from preferably co-sited cells (also labelled with Al, B2, C3 etc. in FIG. 1) and making translation/conclusions about the supported intra-RAT load in these cells.

In particular, FIG. 11 relates to a method in a tele ¬ communications network containing a plurality of cells defining a coverage area containing a plurality of user devices. The plu ¬ rality of cells comprise a first cell B2, C3 as an active cell for a first radio access technology (RAT) and a second cell Al as an inactive cell for the first radio access technology and an active cell for a second radio access technology. Examples of radio access technologies, also abbreviated as RATs, include GSM, UMTS and LTE .

User devices in the at least one first cell B2, C3 are triggered to report measurement information regarding the second radio access technology of the second cell Al . The measurement information regarding the second radio access technology of the second cell Al is received in the telecommunications network (e.g. via the first cell B2 or C3 using the first RAT or via the second cell Al using the second RAT) to determine whether one or more user devices in the at least one first cell B2, C3 are eli ¬ gible for being served by the second cell Al using the first radio access technology.

The first radio access technology is activated in the second cell when the one or more user devices are determined to be eligible for being served by the second cell using the first radio access technology.

In order to improve the accuracy of estimations for the first RAT in the second cell based on measurement information for the second RAT in the second cell, the second cell for the first RAT and second RAT are co-sited.

As illustrated in FIG. 11, the decision to switch on inactive cells may again include a QoS verification. Whereas generally, a QoS verification would appear to be unnecessary in case of activation of a new cell in a network (for coverage, ca ¬ pacity and performance are expected to improve) , QoS

verification may still prove useful to decide which cell to ac ¬ tivate in case of multiple alternatives for activating cells.

Again, two exemplary scenarios will be considered. The first scenario assumes two networks of different RATs, A and B with at least one co-sited base station. The second scenario in ¬ volves a case wherein the base station locations are different.

FIG. 12 depicts a scenario with co-sited networks of RAT A and B . On the first site both cells, denoted 1A and IB , are active, while on the second site only the cell 2A corre ¬ sponding to RAT A is active. The reactivation procedure

considers reactivation of cell 2B for RAT B in order to relieve the overload experienced in the cells 1A and IB .

Assuming that a QoS verification is part of the reacti ¬ vation decision procedure as depicted in FIG. 11, for a number of users currently served by overloaded cells 1A and IB , it may be estimated (i) what QoS the user devices would experience if handed over to a reactivated cell 2B; and (ii) what QoS those user devices remaining in cells 1A or IB would experience.

Considering e.g. LTE or HSPA technologies, for QoS verification (ii) the QoS can e.g. be estimated by a straight ¬ forward recalculation of the sharing factor 1/N, i.e. the time share that each user device is served on the shared channel.

QoS verification (i) may require the estimation of the achievable QoS at potentially reactivated cell 2B, for which no pilot strength information may be available since this cell is inactive. A possible way to estimate this pilot strength is for the candidate users, currently served in cells 1A or IB, to measure the pilot strength (e.g. RSCP (CPICH in UMTS/HSPA) or RSRP (RS in LTE) ) from active cell 2A, which is co-located with inactive cell 2B. The antenna height h A , transmit power P A and carrier frequency F A applied for cell 2A' s pilot signal, as well as the antenna height h B , transmit power P A and carrier frequency F B that will be applied for cell 2B' s pilot once reactivated is known from the configuration set-up. Consequently, the pilot strength P R , A measured for cell 2A can be mapped to a pilot strength P R , B for cell 2B by using a suitable radio propagation model (e.g. the COST Hata model) as follows:

First, the path loss between the user device and cell 2A is calculated from the measured P R , A as PL A = P A - P R , a , where P R , A denotes the received pilot power level and P A denotes the transmitted pilot power level.

Second, the pilot strength P R , B is estimated by applying propagation related corrections on PL A : P R , B = P B - (PL A + a(F A , F B ) + b(h A ,h B )) Here, a(F A , F B ) is the frequency correction factor and b(h A ,h B ) is the antenna height correction factor derived from the corresponding propagation model. It should be noted that for typical co-located deployments of cell A and B (e.g. height difference up to 2 m) the distance from the user device to cell A and B is assumed the same. Additionally, this may result in a negligible correction factor b(h A ,h B ).

In the second scenario for non co-sited cells A and B the corrections on PL A , as presented above, may not be applicable to calculate the pilot received signal strength P R , B . In this case, position information for user devices can be used to obtain the path loss PL B between the user device and the site of cell 2B by using a detailed propagation prediction database

(e.g. predicting the pilot signal strength per location) . Consequently, the pilot strength P R , B is calculated as P R , B = P B - PL B . For user devices equipped with a GPS receiver, the position information of the user device can be obtained using the GPS system. For user devices without a GPS receiver, the position information could still be obtained using e.g. the downlink sig- nals from multiple base stations, for instance as specified in 3GPP TS 36.305 for LTE and TS 25.305 for UMTS/HSPA.

Combined with a measurement of the current interference level on the corresponding carrier, an SINR estimate can be made, which is in turn readily mapped to an estimated bit rate the candidate user device would experience in cell 2B, assuming exclusive access to the cell's shared transport channel. It is noted that the interference level might be different before and after the reactivation of the concerned cell, since there is at least one more cell active after the reactivation procedure.

The decision procedure for cell reactivation typically evaluates different scenarios in terms of the set of user de ¬ vices that is handed over from cells 1A and IB to cell 2B and/or other neighbour candidate cells (once reactivated) , where each such scenario leads to a number of user devices served in each candidate cell and hence a corresponding cell-specific sharing factor 1/N. As depicted in FIG. 12, in different scenarios the same user device might be assigned to different candidate cells 2B or 3B. Multiplying the estimated bit rate for each user device by the sharing factor yields the estimated throughputs. As before, the QoS verification entails a comparison of these throughput estimates with the QoS requirements. The scenario which meets the QoS requirements by reactivating the fewest in ¬ active cells may be selected. If two scenarios by chance need to reactivate the same amount of inactive cells, the one which pro ¬ vides e.g. better QoS or consumes less energy may be selected.