METHOD FOR PLACING A BASE STATION ON STANDBY, COMPUTER PROGRAM PRODUCT, DEVICE, BASE STATION AND CORRESPONDING SIGNAL

Description

1. TECHNICAL DOMAIN

[0001] The field of the invention is that of the implementation of radio communication networks, and in particular of the base stations of such networks.

[0002] The invention relates more particularly to the optimization of the energy efficiency of a base station in a radio communication network.

[0003] The invention has many applications, including, but not limited to, mobile radiocommunication networks, for example cellular networks conforming to the latest or future 3GPP (for "3rd Generation Partnership Project") standards. 2. TECHNOLOGICAL BACKGROUND

[0004] In developed countries, network densification and the proliferation of multifunctional mobile devices ("smartphones") and other connected devices or objects are leading to a sharp increase in energy consumption, not only for the devices themselves but also for base stations (for example, "Node Bs" according to the UMTS standard (Universal Mobile Telecommunications System), "eNBs" (evolved Node Bs) according to the LTE standard (Long Term Evolution), or "gNBs" (next Generation Node Bs) according to the 5G standard currently being standardized). The energy cost of maintaining such a network remains high even during periods of low load, particularly at night.

[0005] Furthermore, emerging countries still suffer from poor coverage and uncertain electricity supply, which can lead to slowdowns or even outages in the service provided by a base station.

[0006] Several approaches have been proposed for managing the energy consumption of the network infrastructure, particularly base stations. For example, on-demand activation of base stations offers a radical solution for reducing energy consumption during periods of low load. However, this results in a prolonged standby mode due to the significant time required to restore connections and perform load switching from one base station to another. Furthermore, it can only be applied to dense networks (in terms of the number of base stations in a given area) that allow a base station to transfer its load to one or more other nodes serving the same geographic area.

[0007] Other solutions have been proposed to address these problems. Examples include techniques for waking the base station via a third-party interface (e.g., a LoRa® system) when the load cannot be transferred to another station, or the implementation of Discontinuous Transmission (DTX) cycles. However, such solutions are only applicable when the base station has no terminals to manage. Therefore, a gNB base station serving only a single terminal (even for a voice service) must remain active and cannot enter a standby state using these known methods.

[0008] It can also be noted that sleep modes have been proposed for terminals, on the one hand by acting on the higher layers, in particular at the application level, and on the other hand, by putting the radio interfaces of the terminals into sleep mode, depending on user traffic or battery constraints.

[0009] However, these techniques are not applicable to base stations that manage multiple users, both in standby and connected modes. Indeed, in a base station, a number of signals must be transmitted regularly and continuously, even if the cell is completely empty. The transmission of signaling, particularly synchronization signals, remains the major obstacle to applying standby modes developed for mobile devices to base stations.

[0010] The 3GPP technical standardization document (“NR Network Energy Requirement”, 3GPP DRAFT, R2-163875 NR NETWORK ENERGY REQUIREMENT, 3RD GENERATION PARTNERSHIP PROJECT (3GPP), ... col. RAN WG2, no. Nanjing, China; 2016-05-23 to 2016-05-27) proposes to minimize signaling signals, or, in the case of time-division multiplexing systems (e.g. GSM, TD-SCDMA, TD-LTE), to prioritize frequency multiplexing of data in order to free up time slots during which the radio parts can be put into standby mode.

[0011] US patent document 2013 / 070635 A1 discloses a method for achieving energy savings in a base station of a mobile communication system (e.g., in an LTE network) by closing a radio frequency channel identified as not carrying multicast traffic.

[0012] US patent document 2015 / 223162 A1 discloses a method for saving energy in a wireless communication network by selectively disabling the transmission of certain control signal symbols (e.g., CRS symbols in LTE) in one or more subframes or downlink frames.

[0013] Energy efficiency remains a key challenge in designing the telecommunications systems, and especially the base stations, of tomorrow. Therefore, there is a need for a solution to optimize the energy consumption of a base station. 1. SUMMARY

[0014] Thus, according to a first aspect, the invention relates to a method for controlling a base station in a radio communication network that multiplexes data in time intervals, each having at least one time portion referred to as the "synchronization" portion and at least one time portion referred to as the "user data" portion, organized into resource blocks distributed in time and frequency. The synchronization time portion includes, among other things, signals intended to maintain the network's functionality, in particular synchronization signals. The user data time portion includes, among other things, data intended for at least one terminal and dedicated signaling (channel estimation, connection establishment messages, etc.).

[0015] According to the invention, the method comprises, for at least some of said time intervals: a step of dividing said user data time portion into at least two sub-portions, an active time sub-portion, suitable for grouping resource blocks to be transmitted, and at least one time sub-portion not containing used resource blocks, called an inactive time sub-portion; a step of putting said base station into standby mode for at least part of the duration, in particular the entire duration (or at least a maximized duration) of said inactive time sub-portions (which may extend over several tens of milliseconds); and a step of modifying the total duration of said time intervals, the number of time slots forming said time intervals being chosen so as to be an integer submultiple of a maximum number NT max< of time slots, the number NT max<of time slots corresponding to a maximum duration between two time portions of synchronization from a given time reference.

[0016] Thus, the invention proposes a novel and inventive solution for minimizing the energy consumption of a base station even under non-zero load conditions, i.e., even when the base station is managing one or more active terminals. The succession of a synchronization time period, an active sub-period, and an inactive sub-period forms a gDTX ("next generation Discontinuous Transmission") cycle.

[0017] To achieve this, the base station's behavior is controlled, and standby periods are introduced after the data to be transmitted has been organized accordingly. The periodic broadcasting of synchronization signals (accompanied, where applicable, by signaling signals dedicated to maintaining the network), as defined by a given standard implemented by the base station in question, can be maintained (5G, for example). However, the transmission of the data portion can be interrupted for a portion of a time interval (a few milliseconds or tens of milliseconds), with the base station entering standby mode and thus conserving energy. This allows a large number of base station components to be shut down, approaching the energy consumption of a base station in standby mode, where only the wake-up functions remain active.

[0018] The periodic broadcasting of network maintenance signaling, in particular synchronization signals, during the synchronization time portion, defines those time intervals (i.e. the duration between two broadcasts of this signaling, which may be variable in some embodiments) during which terminals can be served by the base station.

[0019] Furthermore, the resource blocks used by the various terminals managed by the station are grouped with the resource blocks dedicated to broadcasting synchronization signals. In this way, an inactive time sub-portion is reserved, the duration of which can vary (for example, depending on operational parameters such as a limitation in the base station's energy supply, for example, if it depends on solar or wind power), during which the base station can be put into standby mode in order to reduce its consumption without impacting the overall service obtained in the cell managed by the base station, or at least minimizing such an impact.

[0020] According to a particular embodiment, the process includes a step of distributing a set of resource blocks carrying data from a temporal portion of user data to be emitted in said active sub-portion. This step can notably be implemented by a scheduler, known in itself.

[0021] The active sub-portion can be set large enough to allow the various stakeholders reasonable leeway for the development of their own planning, or "scheduling", algorithms.

[0022] The process includes updating the base station's sleep time accordingly.

[0023] Thus, the duration of the inactive sub-portion is updated, either for implementation directly during the current time interval (i.e., during the time interval that includes the given synchronization portion), or during a subsequent time interval.

[0024] More specifically, the process includes a step to modify the number of time slots. NT forming a time interval.

[0025] Thus, the duration of the time intervals is adaptable, for example to the operational conditions of the base station. This aspect, known in itself in particular in the 3GPP standard, proves to be an effective complement to the general approach of the invention.

[0026] Indeed, we are then dynamically optimizing two values: the number NTof time slots within a time interval, on the one hand, and the number N of time slots during which the base station can be put into standby mode (i.e., the time slots forming the inactive portion of time). The base station therefore only needs to be active during NT - N time slots of the time interval.

[0027] More specifically, during the step of modifying the total duration of said time intervals, the number of time slots forming said time intervals is chosen so as to be an integer submultiple of a maximum number NT max< of time slots.

[0028] Thus, for a duration of a time interval selected from a set of permitted durations (predefined in a standard), a diffusion of synchronization signals at fixed times remains obtained periodically (i.e., at times corresponding to multiples of NT max<). Consequently, a device attempting to synchronize with the base station for very long periods, for example one or more hours, is still guaranteed to be active at the same time as the base station.

[0029] According to a first approach to this particular embodiment, the process includes, after a modification of the number of time slots, a step of adapting the duration of the current sub-portions.

[0030] Thus, to maintain the broadcast of synchronization signals at fixed times (i.e., corresponding to multiples of NT max< ) despite the change in the duration of the time intervals, the duration of a transition interval is adapted, such an adaptation being able to correspond either to an increase or a reduction in the number of time slots.

[0031] According to a second approach to this particular embodiment, the process includes a step of delaying the application of the new duration of the inactive subportion, the modified number of time slots replacing the current number of time slots only q time intervals later, q being determined so as to maintain diffusion of the synchronization signals at fixed times.

[0032] Thus, the periodicity of the emission of the synchronization portion is preserved, via the implementation of a time offset in taking into account the modified duration.

[0033] For example, in a particular implementation, said base station broadcasts said synchronization portions at least every NT max<time slots from a given time reference, and updates at any given instant the number of time slots contained within a time interval. The current number of time slots NT and the modified number of time slots N' T are such that NT max< = k*NT And NT max< = m*N' T , k And m entires and the update takes place during a p -th successive application of the control procedure implementing said current number NT of time slots. The modified value N' T is only applied after q additional time intervals consisting of NT of time slots, and for the K remaining time intervals. K And q are two integers such that kN T = (p+q)NT + KN' T And mN' T = (p+q)NT + KN' T . K = 0 and q = kp if such integers do not exist.

[0034] Depending on the specific implementation, one or more criteria for updating the durations of the time intervals may be taken into account. For example, the determination step and / or the modification step may consider, in particular, at least one operating parameter belonging to the group comprising: a receive switchover request sent by at least one terminal to said base station; a request or end-of-communication signal sent by at least one user terminal to said base station; and an external parameter influencing the broadcasting capacity of said base station.

[0035] Thus, the activity and standby times of the base station take into account the current operational conditions of the base station.

[0036] The number N of time slots forming said inactive sub-portion and / or the number NTforming a time interval is fixed taking into account at least one of the pieces of information belonging to the following group: Guaranteed uptime; type of traffic to be transmitted; availability of electrical power.

[0037] In particular, it is possible to optimize the overall standby time, taking into account the different types of traffic to continuously adapt the values ​​of N and / or NT.

[0038] According to a particular embodiment, an active temporal sub-portion is placed temporally immediately before and / or after a synchronization portion.

[0039] Thus, the base station's standby time is maximized. The base station is active (is "out of standby") during the duration corresponding to the synchronization portion and the active time sub-portion, and inactive (is "standby") during the inactive time sub-portion.

[0040] According to a particular embodiment, when a terminal makes a request to start or end communication, or a transfer request, to the network and this request can be handled by said base station or at least one other base station, the method implements a step of choosing a base station to communicate with said terminal, taking into account the distribution between the active and inactive sub-portions in at least one of said base stations.

[0041] Thus, it is possible to implement global cycle optimization across multiple base stations and / or multiple users.

[0042] The standby periods of the network's base stations can be jointly optimized to achieve overall optimization of the energy consumption of the network in question.

[0043] The invention also relates to a computer program product comprising program code instructions for executing the steps of the method for controlling a base station of a radio communication network as described above.

[0044] Depending on the circumstances, this computer program may reside in a base station, in a controller managing multiple base stations, or be implemented in a virtualized environment. It may also be distributed across these different components.

[0045] The invention also relates to a control device for a base station of a radiocommunication network multiplexing data in time intervals each having at least one time portion of synchronization and at least one time portion of user data organized into resource blocks distributed in time and frequency.Such a device includes, in particular, means for dividing said time portion of user data into at least two sub-portions, an active time sub-portion, capable of grouping blocks of resource to be transmitted, and at least one time sub-portion not containing used resource blocks, called an inactive time sub-portion, means for putting said base station into standby mode, for at least part of the duration of said inactive time sub-portions, and means for modifying the total duration of said time intervals, the number of time slots forming said time intervals being chosen so as to be an integer submultiple of a maximum number. NT max< of time slots, the number NT max< of time slots corresponding to a maximum duration between two time portions of synchronization from a given time reference.

[0046] It should be noted here that, depending on the implementation and development, such a module may include hardware and / or software components. The system may also consist of several distinct hardware and / or software elements, or modules, that interact with each other.

[0047] The invention further relates to a base station of a radiocommunication network characterized in that it comprises such a control module.

[0048] Such a base station can in particular be powered by means of solar and / or wind power generation, the approach of the invention making it possible to optimize consumption, in particular according to the production capacity of these means and / or a corresponding battery charge.

[0049] The invention also relates to a signal emitted by a base station of a radiocommunication network to at least one terminal, multiplexing data in time intervals each having at least one synchronization portion and at least one data portion organized into time- and frequency-distributed resource blocks.

[0050] According to this aspect of the invention, said portion of data comprises, for at least some of said time intervals: at least one active sub-portion, grouping data units intended for at least one terminal in at least one sub-portion of said data portion; and at least one inactive sub-portion, during at least a part of which no transmission takes place.

[0051] The invention also relates to a configuration signal, defining, for a given base station, the active and inactive sub-portions, and which can be exchanged with at least one neighboring base station and / or at least one controller, controlling a set of base stations, and / or at least one user terminal. The configuration can, as appropriate, be performed by the controller or by the base station. 4 LIST OF FIGURES

[0052] Other features and advantages of the invention will become apparent from the following description, given by way of illustrative and non-limiting example, and the accompanying drawings, in which: THE figures 1a and 1b illustrate terminals connected to a base station, as well as the time and frequency multiplexing of data exchanged between the terminals and the base station, obtained by applying a known technique; the figure 2illustrates the steps in a base station control process of the figure 1a according to one embodiment of the invention; the figures 3a and 3b illustrate the time and frequency multiplexing of data exchanged between terminals and the base station of the figure 1a obtained by applying the process of figure 2 ; there figure 4 illustrates an example of the power profile consumed by the base station of the figure 1a when she applies the process of the figure 2 ; THE figures 5a, 5b and 5c illustrate the benefit of broadcasting synchronization signals from the base station of the figure 1a at fixed, periodic moments; the figures 6a and 6b illustrate the application of a time offset before taking into account the modified duration of the time intervals according to an embodiment of the invention in order to maintain the periodic diffusion of the synchronization signals as illustrated in the figures 5a, 5b and 5c ; THE figures 7a, 7b And7c illustrate different strategies for connecting a new terminal to a base station according to different embodiments of the invention; and the figure 8 presents a simplified example of the structure of a module designed to implement the control process of the figure 2 ; THE figures 9a to 9c illustrate an optimization of active sub-portions, in a case implementing two types of data to be transmitted. 5. DETAILED DESCRIPTION OF THE INVENTION 5.1 Glossary

[0053] time interval (120): structural element of the signal emitted by a base station, beginning with a synchronization portion (periodicity of "SS-bursts" for 5G); time portion, hereafter and for simplification also "portion": part, or window, of a time interval, namely: synchronization time portion (150): part of a time interval, for example "SS blocks" according to the 3GPP standard, comprising synchronization signals 130 and where applicable other signaling information and where applicable user data (the "SS-blocks" are not transmitted over the entire time / frequency resources of this window and user data may be scheduled in this window);user data time portion (160), hereafter and for simplicity also referred to as "data portion": second part of a time interval, carrying data intended for the various terminals connected to the base station. This data may include, in particular, data directly related to communication (voice, sounds, images, data in the broadest sense) as well as elements of dedicated signaling (channel estimation, connection establishment, etc.); time sub-portion, hereafter and for simplicity also referred to as "sub-portion": part of a user data time portion, namely: active time sub-portion (300a): part containing the user data to be broadcast; inactive time sub-portion (300b): part not containing user data, during which the base station may be put into standby mode;resource block (140): elementary time-frequency block carrying data intended for a terminal; time slot (170): basic time element of a portion of data, or time fingerprint of a resource block in the time-frequency grid, for example a "TTI" or "time slot" for 5G. 5.2 General principle of the invention

[0054] In all figures in this document, identical elements and steps are designated by the same reference.

[0055] The technique described relates to the control of a base station, or a set of base stations, of a radiocommunication network multiplexing data in time intervals each having at least one synchronization portion and at least one data portion organized into resource blocks distributed in time and frequency.

[0056] More specifically, the general principle of the described technique is based on grouping, for at least some of the time intervals, all the resource blocks carrying data intended for at least one terminal in at least one sub-portion of the data portion, called the active sub-portion, so as to provide in the user data portion at least one sub-portion not containing used resource blocks, called the inactive sub-portion.

[0057] In this way, the base station can be put into standby mode for at least part of the duration of the inactive sub-portion(s), saving energy while limiting the impact on the overall service. In this description, standby mode is maximized and applied over the entire duration of the inactive sub-portion, and the transition and wake-up times that occur during the duration of the inactive sub-portion are considered part of the standby time. According to another interpretation, these transitional phases are not, strictly speaking, part of the standby time.

[0058] The invention thus makes it possible to dynamically optimize the characteristics of time intervals, and consequently the standby time, of a base station, or a set of base stations, in particular to follow network developments, for example the number of sessions, QoS (quality of service) and / or energy constraints.

[0059] Optimization can be local, meaning it's decided independently for each base station (gNB), or centralized, with joint management of multiple gNBs, notably through mechanisms like SON (Self-Organizing Network). Several values ​​of N can also be defined to differentiate traffic, depending on the required QoS and / or user type. For example, the following can be taken into account: a differentiation of "premium" and economy packages; a differentiation of priority data (e.g., medical services, emergency calls) and "best effort" traffic (e.g., downloading a file).

[0060] This is illustrated, for example, on the figures 9a to 9c commented on later.

[0061] The solution of the invention can also be integrated in a "network slicing" context. This concept is a key aspect of 5G, enabling the joint adaptation of RAN and core configurations to achieve a specific QoS objective and thus ensure communication for applications with very different constraints, such as mobile broadband or IoT (Internet of Things). 5.3 Reminder: 5G signal structure

[0062] We now describe, in relation to the figures 1a and 1b ,time and frequency multiplexing, according to a known technique, of data exchanged between 110 terminals and a 100 base station of a radio communication network.

[0063] More specifically, terminals 110 and base station 100 exchange data according to the 5G standard currently being standardized within 3GPP. In other embodiments, terminals 110 and base station 100 exchange data according to other protocols, for example according to the LTE (for "Long Term Evolution") or LTE-Advanced protocol.

[0064] Back to the figure 1Base station 100 broadcasts signals at regular intervals dedicated to maintaining the network, including synchronization signals (e.g., PSS (Primary Synchronization Signal) and SSS (Secondary Synchronization Signal)) and signals indicating the general cell configuration (e.g., PBCH (Physical Broadcast Channel)). Base station 100 also transmits user data signals such as drivers and connection management messages.

[0065] The synchronization signals 130 are grouped into a portion 150 called the synchronization portion. These synchronization portions 150 define time intervals 120, each comprising a synchronization portion 150 and a data portion 160.

[0066] More specifically, according to the prior art, the resource blocks 140 (i.e. the time-frequency blocks) intended for the exchange of data between the terminals 110 and the base station 100 can be distributed throughout the portion 160 of data depending on the implementation of a scheduler and the arrival of this data or the associated signaling.

[0067] Thus, according to known techniques, the base station 100 must remain continuously operational in order to periodically broadcast the synchronization signals 130 in the synchronization portions 150 (which may also contain user data and / or other signaling information), and to receive or transmit the resource blocks 140 carrying the data from the terminals 110 in the data portions 160. According to known techniques, such resource blocks 140 can be assigned to any time slot 170 in the data portion 160. Here, and in the case of 5G, a time slot 170 is understood to be the time footprint of a resource block 140 in the time-frequency multiplexing grid of the blocks in question, as defined by the radiocommunication standard in question. 5.4 Sleep mode

[0068] We now describe, in relation to the figure 2The steps of a method for controlling base station 100 according to different embodiments of the invention. Furthermore, the steps in question, as well as their implementation in the different embodiments considered, are illustrated by examples described in relation to the figures 3a and 3b , 4 , 5a to 5c, 6a and 6b And 7a to 7c .

[0069] Thus, for at least some time intervals 120, during a step E210 optimization of resource use, known as grouping ( figure 2 ), the set of resource blocks 140 carrying data intended for at least one terminal 110 is grouped in at least one sub-portion of the data portion 160, called the active sub-portion 300a, so as to provide in the data portion 160 at least one sub-portion not containing used resource blocks 140, called the inactive sub-portion 300b.

[0070] Thus, the signal emitted by base station 100 includes, for the time interval 120 considered, and more precisely for the time portion of data 160 at least one active sub-portion 300a, grouping the data to be transmitted, and at least one inactive sub-portion 300b, during at least a part of which no transmission takes place.

[0071] Thus, during a step E230 ( figure 2 ), base station 100 is put into standby mode for at least part of the duration of the inactive sub-portion(s) 300b.

[0072] In this way a standby cycle, during which the consumption of base station 100 is reduced (no data transmission), is obtained for a part of the duration of each time interval 120 in question.

[0073] Furthermore, terminals 110 can remain served by base station 100 during such sleep cycles. This avoids a load transfer from base station 100 to another base station in the network.

[0074] In the embodiment illustrated on the figures 3a and 3b , the active subportion 300a is placed temporally immediately before ( figure 3b ) or after ( figure 3a ) a 150 synchronization portion. In this way, the duration of the inactive sub-portion 300b is maximized as well as the corresponding standby time.

[0075] Alternatively, it could comprise two parts, one immediately before and the other immediately after the synchronization portion. In all cases, the synchronization and active portions are contiguous, forming a single duration during which the base station is not in standby mode and maximizing the inactive period during which it is in standby.

[0076] In the embodiment illustrated on the figure 2 , stage E210 Optimizing resource utilization (leading to the grouping of user data) includes: a step E210a of dividing into active and inactive sub-portions, including the determination of a (maximum) number N of consecutive time slots required corresponding to the inactive sub-portion 300b (and consequently a number NT - N of time slots for the active sub-portion 300a); and a step E210b updating, via a given signaling portion, the distribution of resource blocks in the active sub-portion 300a, and consequently of sub-portions 300a, 300b for a current time interval and / or at least one subsequent time interval.

[0077] This distribution operation, as such, is handled by a scheduler. It is adapted according to the duration selected for the active sub-portion. Depending on the implementation, step E210a may or may not be integrated into the scheduler and / or the corresponding scheduling algorithm.

[0078] Thus, depending on the variants considered, the duration of the inactive sub-portion is updated, either for implementation directly during the current time interval (i.e. during the time interval which includes the given synchronization portion 150), or during a subsequent time interval depending, for example, on the operational conditions of base station 100.

[0079] In some variations, such operational conditions may be linked to operating parameters belonging to the group comprising: a receive handover request sent by at least one 110 terminal to base station 100; a call start or end request signal sent by at least one user 110 terminal to base station 100; and an external parameter influencing the broadcasting capacity of base station 100 (e.g., weather forecasts for a base station powered by solar panels).

[0080] The standby step E230 (according to any of the aforementioned embodiments) is looped back to the regrouping step E210 (according to any of the aforementioned embodiments). In this way, the base station control process is applied successively to a series of time intervals 120 so as to track the evolution over time of the base station's operational conditions 100.

[0081] In some embodiments, this evolution can be taken into account remotely, by a controller controlling a plurality of neighboring base stations, and / or by remote control means, housed for example in a "cloud" and / or a management center (data centers).

[0082] It is indeed noted that network architectures are converging towards "virtualization," meaning that the base station as we know it today could, in the future, essentially resemble a simple antenna, with most functions (related, for example, to connection management and QoS) being relocated to data centers. Ultimately, we will have a "virtual" base station that is "physically distributed" into RRHs (remote radio heads) and BBUs (baseband units).

[0083] Thus, the process, device or base station can consist of several distributed elements, implemented in hardware and / or software. 5.5 Illustration of energy savings

[0084] More specifically, the amount of energy saved during a standby period is not simply linear with the standby duration. It appears that longer standby periods allow for more economical operating modes by enabling the suspension of power to components requiring long periods of shutdown and / or wake-up, as illustrated in the... figure 4 .

[0085] In practice, the operating power of the components (RF module, decoder, etc.) of base station 100 can vary between 490W (signaling only) and 750W (full load). However, the switching off and waking up of these components are not instantaneous, and the standby capabilities are severely limited by the transition time (deactivation time + activation time).

[0086] Analysis of base station consumption has allowed us to define 4 standby modes (see. « A Flexible and Future-Proof Power Model for Cellular Base Stations”, Björn Debaillie, Claude Desset, Filip Louagie, IEEE VTC Spring 2015 ), or SM (for "Sleep Mode" in English): a) SM1: This mode primarily involves disabling the power amplifier, and its transition time is only 71 µs (≈1 OFDM symbol). The operating power of a base station in this standby mode is estimated at 157W. b) SM2: With a transition time of 1 ms, the operating power is reduced to 42.9W. c) SM3: In this standby mode, most components are disabled. Requiring a 10ms transition, this mode limits the power to 28.5W. d) SM4: This mode almost completely shuts down the base station and requires a 1s transition. Only the wake-up functions on the backhaul network remain active, and the operating power is 24.3W.

[0087] As illustrated on the figure 4 For a gDTX 21 cycle of 10ms and an active period 211 of 5ms, the gNB can reach SM2 (212) and the average power consumed is, for example, equal to: 750 W × 5 ms + 42 , 9 W * 4 ms + 750 − 42.9 / 2 W * 1 ms / 10 ms = 431.8 W (the last term corresponds to energy consumption during the transition period).

[0088] For a gDTX cycle of 40ms and an active period of 20ms, the gNB can switch to SM2 (222), then SM3 (223), and the average power consumption is, for example, equal to: 750 W × 20 ms + 28 , 5 W * 9 ms + 750 − 42.9 / 2 W * ms + 49 , 2 − 28 , 5 / 2 W * 10 ms / 40 ms = 392 , 8 W 5.6 Variation in the duration of time intervals

[0089] As planned in the 5G standard currently under development, during a step E220 ( figure 2 ), the duration of the time intervals 120 can be modified so as to be, for example, also adaptable to the aforementioned operational conditions of base station 100.

[0090] Such flexibility makes it possible to optimize, in particular, the duration of the 300b inactive sub-portions and therefore the overall energy consumption of the base station 100. In other words, the invention proposes, according to this approach, to optimize, on the one hand, the duration of the active sub-portion, by limiting it as much as possible (to NT - N time slots), and on the other hand the duration of the inactive sub-portion, by increasing it ( NT time slots for the time interval). The implementation of the invention relies, according to this embodiment, on a pair of values ​​(N, NT ) à to optimise.

[0091] In other embodiments not illustrated on the figure 2 , the active and inactive sub-portions 300a, 300b and / or the duration of the time intervals 120 are however fixed in order to simplify network management.

[0092] Extending the duration of the time intervals 120 allows indeed the transition to particularly economical standby modes (SM2, SM3 or even SM4), and provides substantial energy gains (cf. “Advanced Sleep Modes and their impact on flow-level performance of 5G networks », Fatma Ezzahra Salem, Azeddine Gati, Zwi Altman, Tijani Chahed).

[0093] Base station 100 can announce the organization of its intervals (values N And NT ) and its standby cycle to nearby 110 terminals. In this way, the 110 terminals in question can synchronize their own standby cycles, and in particular their DRX cycles. This announcement can be conveyed, for example, via a SIB message (for "System Information Blocks"). Thus, energy savings are achieved for both the base station and the 110 terminals. This results in an intermittent cell where the 110 terminals and the 100 base station alternate between periods of activity and inactivity. Such a cell is therefore only visible during the 150 synchronization segments and the 300a active sub-segments.

[0094] Depending on the method of implementation of the figure 2 The E220 step for modifying the duration of time intervals 120 defines, for the intervals in question, a number of time slots that is an integer submultiple of a maximum number NT max<of time slots (For example, in the 5G standard, PSS, SSS and PBCH signals are grouped into blocks which are transmitted with a reconfigurable periodicity among the values: 5, 10, 20, 40 ms...).

[0095] Thus, regardless of the duration of the time intervals 120 selected from the permitted durations, a broadcast of the synchronization signals 130 at fixed and predetermined times 500 can be obtained periodically (i.e., at least every NT max< time slots from a given time reference).

[0096] For example, as illustrated on the figures 5b and 5c, such a broadcast of the 130 synchronization signals at the fixed 500 times in question makes it possible to coincide the active period of a terminal coming out of an extended standby period (for example an eDRX cycle of one or more hours) with an active portion of the base station time interval, even if the duration of this interval has been modified.

[0097] However, as illustrated on the figure 6a , it is possible that a modification 61 of the duration of the time intervals may result in a shift 62 in the periodicity of the broadcast of the synchronization signals even when the current duration and the modified duration are selected from the aforementioned authorized durations.

[0098] Thus, in order to maintain the broadcast of synchronization signals at fixed times 500, step E220, which modifies the duration of the time intervals 120, includes a step E220a ( figure 2) adaptation of the duration of the current inactive subportion 300b (i.e. either an increase or a reduction in the duration of the inactive subportion).

[0099] Following the example illustrated on the figures 6a and 6b , the introduction of an additional correction period, T offset , for a so-called transition time interval, allows the broadcasting of 63 synchronization signals to be maintained at the fixed times 500 in question despite the change in the duration of the time intervals 120 from the current duration of NT time slots towards the modified duration of N' T time slots.

[0100] In other unillustrated examples, the preservation of the diffusion of synchronization signals at fixed times 500 can be achieved by reducing the duration of the inactive subportion.

[0101] In other embodiments not illustrated on the figure 2The E220 step for modifying the duration of the time intervals 120 does not include an E220a step for adapting the duration of the current inactive sub-portion 300b. Instead, in order to maintain the broadcast of the synchronization signals at the aforementioned fixed times 500, the modified number N' T time slots replace the current number NT of time slots after a given number q of additional applications of the control process implementing the current number NT of time slots to a series of q successive time intervals 120 consecutive to the current time interval 120 (i.e. consecutive to the time interval 120 during which the modified number N' T is obtained).

[0102] Thus, the diffusion of synchronization signals at the aforementioned fixed times 500 is maintained simply through the implementation of additional iterations of the method according to the invention based on the current number NT of time slots.

[0103] For example, in a particular implementation, said base station broadcasts said synchronization portions at least every NT max< time slots from a given time reference, and updates at any given instant the number of time slots contained within a time interval. The current number of time slots NT and the modified number of time slots N' T are such that NT max <= k*NT And NT max< = m*N' T , k and m integers and the update takes place during a p -th successive application of the control procedure implementing said current number NT of time slots. The modified value N' Tis only applied after q additional time intervals consisting of NT of time slots, and for the K remaining time intervals. K and q are two integers such that kN T = (p+q)NT + KN' T And mN' T = (p+q)NT + KN' T . K = 0 and q = kp if such integers do not exist. 5.7 optimization taking into account at least one other nearby base station

[0104] Back to the figure 2 , and in connection with the figures 7a to 7c , when a 110' terminal, for example, makes a request to the network to initiate or terminate communication, or to transfer to another base station, and this connection can be handled by base station 100 or at least one other 100' base station, the method according to the invention further comprises a step E200 selection of a base station for terminal 100', which takes into account in particular the distribution between the active sub-portions 300a and inactive 300b in at least one of said base stations 100, 100'.

[0105] Thus, a centralized optimization of the standby cycles of base stations 100 and 100' makes it possible to combine the management of the cycles in question while taking into account the possible load transfers from one base station to another (" traffic offloading " in English).

[0106] For example, as illustrated on the figure 7b , terminal 110' can be transferred to base station 100, which is already reasonably active, in order to leave base station 100' in an advanced standby state, for example in SM2, SM3 or SM4.

[0107] Conversely, as illustrated on the figure 7c, if additional traffic prevents base station 100 from implementing a grouping of resource blocks 140 so as to obtain an inactive subportion 300b in order to maintain itself in a standby state, it may be preferable to let terminal 110' be served by base station 100' in order to obtain a standby cycle for each of the two base stations 100, 100' (see the respective standby cycle profiles gDTX1 and gDTX2).

[0108] As already mentioned, the corresponding decisions can be made by a controller controlling multiple base stations and / or in a virtualized way, for example via a data center. 5.8 Possibility of adaptation according to the type of traffic

[0109] According to one aspect of the invention, it is possible to adapt, or rather differentiate, the values N And NTdepending on the type of traffic. By traffic type, we can mean differentiation based on the required quality of service, or QCI (QoS Class Identifier - 4G / 5G), by network slicing, but also, for example, by differentiating user data from signaling (e.g., ACKs). A base station that handles several types of traffic therefore adapts its gDTX cycle according to each use case, and not according to the most demanding use case.

[0110] An example of adaptation is illustrated by the figures 9a to 9b . There figure 9a illustrates two distinct types of traffic, the first (91) requiring low latency (high periodicity) but few resources, the second (92) requiring more resources (i.e. a smaller N), but with a less strong delay constraint.

[0111] To transmit all the data corresponding to both traffic flows, a direct approach (without differentiation) would lead to the situation illustrated in the figure 9b : we retain the values ​​of N and NT the weakest, which leads to reduced standby time. On the other hand, by taking into account the differentiation of content, it is possible to optimize its transmission, as illustrated in the figure 9c , to obtain much longer standby times. For example, we retain the value NT of the first type of traffic (the weakest), but we transmit alternately: active sub-portions 93, containing data relating to both traffic 91 and 92; and active sub-portions 94, containing only data relating to traffic 91.

[0112] Thus, we respect the QoS (quality of service) required for each type of traffic, while increasing standby times. 5.9 Implementation of the process in a technical device

[0113] There figure 8 schematically presents an example of an 800 module for controlling base station 100. Such an 800 module enables the implementation of the control process of the figure 2 The module 800 here comprises a random access memory 803 (for example, RAM), a processing unit 802 equipped, for example, with a processor, and controlled by a computer program stored in read-only memory 801 (for example, ROM or a hard drive). At initialization, the computer program's code instructions are, for example, loaded into the random access memory 803 before being executed by the processor of the processing unit 802.

[0114] This figure 8 illustrates only one particular way, among several possible ways, of implementing the technical means included in module 800, so that it performs at least certain steps of the control process detailed above, for example in relation to the figure 2(in any of the various embodiments). Indeed, these steps can be performed interchangeably on a reprogrammable computing machine (a computer, a processor, or a microcontroller) executing a program comprising a sequence of instructions, or on a dedicated computing machine (for example, a set of logic gates or any other hardware module). If the means included in module 800 are implemented with a reprogrammable computing machine, the corresponding program (i.e., the sequence of instructions) can be stored on a removable or non-removable storage medium, this storage medium being partially or fully readable by a computer or a processor. It can also be downloaded via any suitable transmission method.

[0115] In some embodiments, such an 800 module is included in the base station 100.

[0116] In other embodiments, the module according to the invention may be embedded software, for example in the scheduler, the overall base station control software, or be dedicated software. The module may comprise hardware and / or software components.

[0117] Furthermore, the invention can be implemented, in part, outside of a base station. Specifically, the operations for determining the durations of the active and inactive sub-portions can be performed, as appropriate, by the base station or by a controller managing several neighboring base stations. This controller optimizes the durations of the active and inactive sub-portions for each base station, taking into account various factors (e.g., the energy reserves of each cell). The method, the computer program product, and / or the module can also be distributed between this controller and the base stations. 5.10 Compatibility with DRX (discontinuous reception) of UEs (user equipment) and management of IoT (Internet of Things) devices

[0118] The proposed solution does not interfere with DRX, eDRX, and PSM (power-saving mode) cycles, used particularly for IoT devices. Indeed: 1) Unlike previously proposed techniques (particularly in the context of IoT), the gNB does not adapt its gDTX standby cycle to the DRX cycles of the UEs, but rather announces its gDTX cycle to the UEs to allow them to synchronize their DRX cycles. Thus, the implementation of the gDTX cycle ensures energy savings for both the gNB and the UEs. 2) DRX / gDTX synchronization can be achieved by adjusting the DRX cycle parameters, as defined by the 3GPP standard. For example, for a UE in RRC_CONNECTED mode, the "OnDuration" timer is set to a maximum of NT - N time slots, and the value for "DRXShortCycle" is configured to NT. The "DRX-InactivityTimer" can be set to zero so that the UE immediately starts its DRX cycle. At the end of a "DRXShortCycleTimer" timeout (i.e., if no data has been received), the UE switches to a long cycle which can be configured in the same way.3) eDRX cycles (“extended DRX” in English) and PSM (“Power saving mode” in English) are used mainly for loc devices and can be an obstacle to parameter optimization. N And NT Indeed, an eDRX or PSM cycle for an IoT device can last several hours. It is therefore possible that the gNB has altered the time interval. NT during the standby period of the IoT device, and consequently, that this device wakes up during the inactive sub-portion of the gNB. 4) By guaranteeing an activity period on a fixed basis of N T max time slots (e.g., 80ms), the invention supports the implementation of eDRX and PSM cycles. It is sufficient that the corresponding timers, as well as the TAU (Tracking Area Update), are synchronized to the guaranteed activity periods, i.e., that they are equal to a multiple of N T max .Thus, a UE (smartphone or IoT device) can enter a standby state to n × N T max time slots, Vn EN, and be assured of finding the gNB active upon waking, regardless of any changes made to NT or N. In the specific case of the PSM, the arrival of uplink data also triggers the UE to wake up. It is then subject only to a maximum additional delay of N T max time slots. 5) A time offset (in English, T officer ) can be used if the parameter is modified NT. This is equivalent, for example, to delaying the application of a new value by a few cycles. NT in order to guarantee an active sub-portion every N T max time slots. This offset can also be used to facilitate the transition between the announcement of a new cycle and its effective implementation in the EUs.

Claims

1. Method for controlling a base station (100) of a radiocommunication network multiplexing data in time intervals (120) each having at least one synchronization time portion (150) and at least one time portion (160) of user data which are organized in resource blocks (140) distributed in time and in frequency, a time interval being formed of time slots, a time slot corresponding to a time print of a resource block distributed in time and in frequency, characterized in that it comprises, for at least some of said time intervals (120): - a step of dividing (E210a) said time portion of user data into at least two subportions, an active time subportion (300a), which is able to group resource blocks to be transmitted, and at least one time subportion not containing used resource blocks, referred to as an inactive time subportion (300b), - a step of putting said base station on standby (E230) for at least part of the duration of said one or more inactive time subportions, and - a step of modifying (E220) the total duration of said time intervals, the number of time slots forming said time intervals being chosen so as to be an integer submultiple of a maximum number NTmax of time slots, the number NTmax of time slots corresponding to a maximum duration between two synchronization time portions starting from a given time reference.

2. Method according to Claim 1, characterized in that an active time subportion (300a) is placed temporally immediately before and / or after a synchronization portion (150).

3. Method according to Claim 1, characterized in that the number N of time slots forming said inactive subportion (300b) and / or the number NT forming a time interval is set taking account of at least one of the items of information belonging to the following group: - guaranteed time in service; - type of traffic to be transmitted; - availability of electrical energy.

4. Method according to Claim 1, characterized in that it comprises, after a modification of the number of time slots NT forming a time interval, a step of adapting (E220a) the duration of a current inactive subportion.

5. Method according to Claim 1, characterized in that it comprises, after a modification of the number of time slots NT forming a time interval, a step of delaying the application of the new duration of the inactive subportion, the modified number of time slots being substituted for the current number of time slots only q time intervals later, q being determined so as to keep broadcasting the synchronization signals at set times.

6. Method according to Claim 1, characterized in that, when a terminal can be taken charge of by said base station or at least one other base station, the method implements a step of choosing a base station to communicate with said terminal, taking account of the distribution between the active subportion and inactive subportion in at least one of said base stations.

7. Computer program product comprising program code instructions for executing the steps of the method for controlling a base station of a radiocommunication network according to any one of Claims 1 to 6 when said program is executed by a microprocessor.

8. Device for controlling a base station of a radiocommunication network multiplexing data in time intervals each having at least one synchronization time portion and at least one time portion of user data which are organized in resource blocks distributed in time and in frequency, a time interval being formed of time slots, a time slot corresponding to a time print of a resource block distributed in time and in frequency, characterized in that it comprises means for dividing said time portion of user data into at least two subportions, an active time subportion, which is able to group resource blocks to be transmitted, and at least one time subportion not containing used resource blocks, referred to as an inactive time subportion, means for putting said base station on standby for at least part of the duration of said one or more inactive time subportions, and means for modifying the total duration of said time intervals, the number of time slots forming said time intervals being chosen so as to be an integer submultiple of a maximum number NTmax of time slots, the number NTmax of time slots corresponding to a maximum duration between two synchronization time portions starting from a given time reference.

9. Base station or group of base stations of a radiocommunication network, characterized in that it comprises at least one device according to Claim 8.