FR2985151A1 - METHOD FOR CONFIGURING STATIONS IN A WIRELESS COMMUNICATIONS NETWORK - Google Patents

Abstract

The invention relates to a method for configuring stations of a wireless communications network, said stations receiving, according to a predetermined protocol, the results of measurements of interference from client devices to which they are connected, said method comprising: next step: a) said stations transmit (E1, E2) data representative of said measurement results to a central management entity to which they are connected. Said method is remarkable in that it further comprises the following steps: b) said management entity aggregates (E3) said data to construct a global interference graph, c) the management entity segments (E4) the graph in station communities, d) performing (E5, E6, E7) an optimization calculation for each community, and e) each station applies (E8) the result of the optimization calculation which concerns it. Application particularly to femto or pico networks.

Description

The invention relates to the field of telecommunications. More particularly, the invention relates to the minimization of interference between neighboring wireless stations, these stations being able to operate in the radio frequency domain or in another frequency domain (for example the infrared). The invention relates in particular to the problem of "off-loading" (in English), in which radio stations, called "unloading" stations, treat a part of the communications traffic so as to unload other stations and reduce congestion in high-load areas, especially at times of the day when communication traffic is high. The invention is particularly compatible with all known radio access technologies (TDMA, CDMA, W-CDMA, OFDMA, and so on). It applies in particular to cellular networks using GSM / GPRS technology, as defined in version 97 and later versions of the GSM standard, or UMTS ("Universal Mobile Telecommunications System") technology, as defined in particular in the 23.002, 23.003 and 29.060 of the 3GPP project ("Third-Generation Partnership Project"), the LTE ("Long Term Evolution"), or the HSPA ("High Speed Packet Access") standard. It is recalled that, while the GSM uses a circuit-switched data transmission medium ("circuit-switched"), the GPRS adds a new packet-switched data transmission medium ("packetswitched"). in English), which provides a mobile station with permanently available Internet Protocol (IP) connectivity, but in which radio resources are allocated only when data is to be transferred. Subscribers of a mobile operator can thus access services using the IP protocol, such as e-mail, file downloads, website consultation or WAP. UMTS uses circuit mode switching and packet switching. UMTS uses W-CDMA technology, standardized by 3GPP, and is the European implementation of ITU's IMT-2000 specifications ("International Telecommunication Union") for 3G cellular radio systems. UMTS makes it possible to exchange data contained in IP packets with servers belonging to a network outside the UMTS network, such as the Internet network. The LTE standard is part of the UMTS standard, but incorporates many modifications and enhancements, including the use of OFDM modulation for the downlink and SC-FDMA modulation for the uplink (instead of W-CDMA for UMTS). LTE requires dedicated radio coverage, distinct from UMTS coverage. HSPA is the combination of HSDPA (High Speed Downlink Packet Access) for the downstream, and HSUPA (High Speed Uptink Packet Access) for the uplink. The explosion of multimedia data traffic in cellular mobile networks is a new challenge for operators, especially in high traffic areas. A known solution for increasing the capacity of a network is to set up additional stations ("hot spots" in English) in high traffic areas. In particular, cells managing cells of relatively small or very small size are used in addition to "macro" cells of conventional size. A cell covering a zone having a width of less than two kilometers is generally called a "microcell", a "picocell" a cell covering an area of about 200 meters, and a "femtocell" a cell covering a zone of the order. 10 meters. The coverage area of each station can be limited by controlling its power.

Indeed, the flexibility in terms of cell size is a feature of 2G mobile technology and the following technologies, and is a significant factor in increasing network capacity. The power controls implemented in mobile networks make it easier to reduce interference between neighboring cells using the same frequencies. By subdividing cells and creating more cells (referred to as network densification) to serve high density areas, a cellular network operator can optimize spectrum utilization and increase capacity.

It is thus possible, by means of a low-power base station, to cover a limited area. This makes it possible to increase the capacity of a network in areas of difficult or expensive access in the conventional "macrocell" approach. Specifically, a microcell can cover an area such as a shopping center, a hotel, or a transportation center. Microcells are often deployed temporarily during sporting events and other occasions where it is known in advance that additional capacity will be needed at a specific location.

A picocell usually covers a small area, for example inside a building (offices, shopping malls), or, more recently, inside an airplane. Picocells are typically used to extend the coverage to the inner areas, that is, where the external signals do not penetrate well, or to increase the network capacity in areas of high density of phone use, such as in the train stations. A picocell base station is usually a low cost, small size, and relatively simple unit. Picocells are found for most cellular technologies such as GSM / GPRS, UMTS and LTE.

In GSM / GPRS networks, in particular, each picocell base station is connected to a BSC (initials of the words "Base station Controller" meaning "Base Station Controller"). In fact, many picocells are connected to each BSC; the BSC manages radio resources and handover functions (migration of a mobile terminal from one cell to a neighboring cell), and aggregates data for transmission to the Mobile Switching Center (in English, "Mobile Switching Center" (or MSC) and / or the Gateway GPRS Support Node (GGS). Connectivity between the picocell units and the BSC is generally accomplished by means of inside wiring to a building; the latest systems use Ethernet cabling; planes use satellite links. The most recent developments concern a unit containing not only a picocell station, but also capable of performing many functions of the BSC and some of the MSC. This type of picocell is sometimes called an "access point base station" or a "corporate femtocell". In this case, the unit contains all the means required to connect directly to the Internet, without the need for the BSC / MSC infrastructure. This is potentially a more profitable approach. As for UMTS networks, they may include private networks (for example home networks) consisting of small cells called "femto NodeB", in which the base stations, called "Home NodeB" (HNB), each combine the functions of NodeB and RNC (initials of the words "Radio Network Controller" meaning "Controller of Radio Resources"). Each HNB is connected to a gateway ("HNB Gateway") located outside the private network in the operator's radio access network; the HNB Gateway manages HNB and subscriber traffic, and mediates with the core network. Note that GSM networks also include HNB Gateway gateways with similar functionality. In addition to the technologies described above, the invention applies in particular to "femto 3G" architectures as defined in document TR 25.820 V8.2.0 (2008-09) of the 3GPP, as well as to future "femto LTE" architectures. ". These architectures use femtocells in which a low cost mobile home network can be deployed using the broadband infrastructure already present at the subscriber; the "backhaul" (link between the stations and the core network) is provided by a wired connection, for example an ADSL connection. This context is of the utmost interest to the operators: in fact, recent studies have shown that telephone calls mainly take place in the user's home where the environment is not conducive to good propagation of the waves from the stations. macro; this results in a greater consumption of radio resources because of the automatic adjustment mechanisms of the MCS (initials of the words "Modulation and Coding Schemes" meaning "Modulation and Coding Schemes") in the stations, a level of interference more important because of the automatic power control, as well as a possible degradation of the user flow and, ultimately, of the QoS (initials of the English words "Quality of Service" meaning "Quality of Service"). It should also be noted that femtocells will become essential in buildings complying with the new HQE (High Environmental Quality) standard, as the power loss of the radio signal crossing the walls of these buildings may exceed 20 dB. It is therefore expected that a significant proportion of future telephone offers will include a femtocell. In addition, these femtocells will be able to open their access to other users (open access or hybrid access). The invention therefore preferably relates, but not exclusively, stations managing "pico" or "femto" type cells, used in a domestic environment or within a company. If the idea of a massive deployment of such stations is attractive, its implementation faces many difficulties. Indeed, the number of sites considered (one per household, and per company) and their location in the home of the user or in a company, make it impossible to plan deployment. Thus, it is inevitable that some sites, once deployed, interfere strongly without it being possible to adjust their powers manually, like domestic WiFi access. Domestic stations are used to reduce the load on macro networks and must therefore provide a higher or equal level of QoS by reusing the narrow licensed spectrum bands of the macro network. These two constraints (QoS and narrow band) create an even more demanding context than is currently the WiFi. In particular, it is difficult for an operator to manage interference in a network in perpetual evolution, and consists of millions of nodes whose deployment situation is unknown to the operator since this deployment is performed by subscribers. The present invention thus relates, according to a first aspect, to a method of configuring stations of a wireless communications network, said stations receiving according to a predetermined protocol the results of interference measurements on the part of client devices to which they are connected, said method comprising the following step: a) said stations transmit data representative of said measurement results to a central management entity to which they are connected. Said method is remarkable in that it further comprises the following steps: b) said management entity aggregates said data to construct a global interference graph, c) the management entity segments the graph into communities of stations, d. ) an optimization calculation is carried out for each community, and e) each station applies the result of the optimization calculation which concerns it. Thus, the method according to the invention makes it possible to adjust the power levels of the stations so as to limit the impact of mutual interference. Advantageously, this method is automatic and very well adapted to an implementation within a set comprising a very large number of stations. The proposed solution makes it possible to efficiently discharge the macro-cellular network from communications carried out at the home of the users or within a company. The QoS of the users attached to the stations is also improved because the proposed solution allows an autonomous planning of the frequencies within the network of the stations. Thus users enjoy stable performance regardless of the activity of neighboring stations. Finally, the partially centralized nature of the solution provides the operator with a means of supervising his network of stations, and thus facilitates the detection of faults. The invention applies to cellular mobile networks, but also to wireless networks (radio frequencies or other), such as Wi-Fi networks, in which it is possible to set up a transmission to the stations of interference measurements carried out by terminals. According to particular features: in said step a), each of said stations aggregates said received results to obtain a joint distribution of interference in the form of an interference subgraph of which this station is the center, then sends its sub-graph interference graph to said central management entity, and in said step b), the management entity assembles the received interference sub-graphs to construct said global interference graph. Thanks to these provisions, the tasks leading to the construction of the global interference graph are shared between the stations and the management entity.

According to other particular characteristics, during said step d): the management entity informs each station of the structure of its community and of a distributed calculation which is incumbent on it, each station performs an optimization calculation, and each station disseminates the configuration of stations obtained following its calculation to the other members of its community if this configuration is more efficient than a previous or predetermined configuration, and in said step e), each station applies the configuration of stations the most powerful among the set comprising the configuration obtained and the configurations received. Thanks to these provisions, the optimization calculation is distributed between the stations of the same community (rather than being entrusted to the management entity or to an associated computing center), in order to increase the calculation speed. while reducing the amount of signaling in the network.

According to a second aspect, the invention relates to a station configuration device of a wireless communications network, said station being able to receive, according to a predetermined protocol, the results of interference measurements on the part of client devices to which she is connected. Said device is remarkable in that it comprises means for: - aggregating said received results to obtain a joint distribution of interference in the form of an interference sub-graph of which this station is the center, and - to send said sub-graph interference graph to a central management entity to which it is connected. According to particular features, said device further comprises means for: receiving from a central management entity to which said station is connected information concerning the structure of a community of stations to which the station belongs and a distributed calculation that is incumbent on it, - perform an optimization calculation, - disseminate the configuration of stations obtained following its calculation to the other members of said community if this configuration is more efficient than a previous or predetermined configuration, and - apply the configuration of the most powerful stations among the set comprising the configuration obtained and the configurations received. Note that it is possible to realize these station configuration devices in the context of software instructions and / or in the context of electronic circuits. According to a third aspect, the invention relates to a central management entity in a wireless communications network, said management entity comprising means for receiving from the stations to which it is connected data representative of a shared distribution of data. interference associated with each of these stations. Said management entity is remarkable in that it further comprises means for: - aggregating said received data to construct a global interference graph, and - segmenting the graph into communities of stations. According to particular features, said data received from each of said stations being in the form of an interference sub-graph of which this station is the center, it further comprises means for assembling said sub-graphs of interference to construct said global interference graph. According to other particular characteristics, said management entity further comprises means for informing each of said stations of the structure of its community and of a distributed calculation which is incumbent on this station. Note that it is possible to realize these management entities in the context of software instructions and / or in the context of electronic circuits. The advantages offered by these devices and management entities are essentially the same as those offered by the correlative methods briefly described above. The invention also relates to a computer program 25 downloadable from a communication network and / or stored on a computer readable medium and / or executable by a microprocessor. This computer program is notable in that it includes instructions for performing the steps of any of the station configuration methods succinctly set forth above, when run on a computer.

The advantages offered by this computer program are essentially the same as those offered by said methods. Other aspects and advantages of the invention will appear on reading the detailed description below of particular embodiments, given by way of non-limiting examples. The description refers to the accompanying drawings, in which: FIG. 1 schematically illustrates the representation of a set of interfering stations in the form of a graph; FIG. 2 illustrates the steps of a configuration method according to a embodiment of the invention, and - Figure 3 illustrates certain parameters of the calculation of optimization of the transmission power of the stations, according to one embodiment of the invention. Figure 1 shows a set of wireless stations (referred to as "HeNB" in Figures 1 and 2). These stations are considered here as the nodes of a graph. Some pairs of stations are connected by an arc to which has been indicated a measure of the propensity that the two stations connected by this arc to interfere. Due to the geographical distribution of dwellings, this graph, which can be of the order of magnitude of one million nodes, has a community structure that tends to mesh more densely with the nodes inside the same building. a set of adjoining buildings, a subdivision, and so on. It is naturally within these communities (or "clusters" in English) that the interference phenomena are the most troublesome. According to the present embodiment of the invention, the station graph is obtained, and it is divided into communities of predetermined maximum size so that a distributed optimization algorithm can be calculated by the stations in each community.

A wireless network comprising the following elements is considered: - UEs (initials of the words "User Equipment" meaning "User Equipment"), - wireless stations, and - a central management entity for these stations .

It is assumed that, in known manner, each UE participating in the implementation of the invention (at least one of them is required for each station) has means for: - implementing a protocol in which it carries out measurements of field received on a reference channel from its server station and neighboring stations, and - transmit the results of these measurements to its server station. Optionally, said UE furthermore has means for storing the results of these measurements before transmitting them to its waiter station. According to the present embodiment of the invention, illustrated in FIG. 2, the following steps are implemented. Initially, either the stations are in a configuration resulting from a previous application of the invention, or a default configuration is applied. During a step E1, each station aggregates the measurement results received from the UEs, to obtain a joint distribution of interference in the form of a subgraph ("subgraph") of interferences of which this station is the center (step marked "statistical aggregation" in Figure 2). Measurements reported by the EUs must measure the degree of interference between stations. However it is important that this measure integrates the notion of inconvenience by incorporating a notion of frequency of occurrence. Thus, the interference level of a station can be high but very rare, while that of another station can be average but permanent. In order to take this aspect into account, it is possible to ask the UEs to carry out periodic measurements, during their activity phases, of the reference signals coming from the active neighboring stations. At the level of each station, this collection of measurements makes it possible to estimate the joint distribution of its power level and the interference generated by neighboring stations. For a station of index 0 surrounded by n detectable neighboring stations, denote by Do (io, ..., in) this joint distribution, where io represents the power level of the station of index 0, and ik, where k = 1,, n, represents the power level (interference) coming from the next k-th.

From this joint distribution, it is easy to calculate the mathematical expectation of the ratio between the interference level emanating from a neighboring station and the useful signal, namely: E = (111 io D 0 (i 0,, in ) pio ... pin.to We can use this metric (or some other analog) to build the interference subgraphs, but of course these estimators evolve over time, and we need to put in place an implementation strategy. During a step E2, each station sends its interference sub-graph to said central management entity (step marked "subgraph report" in Figure 2). management entity assembles the subgraphs to build a global interference graph (step marked "data aggregation" in Figure 2) .In general, the assembly of the subgraphs gives rise to a directed graph, because the interferences are not necessarily reciprocal, as illustrated by the links between station 1 and station 2 in FIG. 2. However, to simplify the implementation of the invention, one can be content with an approximation in which one (1) replaces this graph oriented by a non-graph. oriented. Here, as examples, two methods to realize this approximation: - we take, as weight of the link between two nodes, the maximum value of the two oriented links; this method tends to protect as much as possible the nodes most affected by the interference; or - we take, as weight of the link between two nodes, the sum of the two links; this method means that interference is considered less serious if one of the two stations experiences less interference.

During a step E4, the management entity segments the graph into communities (step marked "clustering" in Figure 2). This community segmentation requires the choice of a maximum size N, which is limited by the complexity of the distributed optimization algorithm (see below) that will be applied in each of these communities. In addition, the number of sections in the graph must be minimized, so that the stations in severe interference have their powers optimized jointly. Segmentation of such a graph is a difficult problem, but there are powerful heuristics that give good results. Especially the so-called "spectral" techniques (see U. von Luxburg's article "A Tutorial on Spectral Clustering", Statistics and Computing, 17 (4), 2007), and so-called "greedy" techniques ( see the article by V. Blondel, J.-L. Guillaume, R. Lambiotte and E. Lefebvre entitled "Fast Unfolding of Communities in Large Networks", Journal of Statistical Mechanics: Theory and Experiment, 2008). In this last document, the generated communities do not have a size limit, but it is possible to modify the proposed algorithm to impose a maximum size N, by performing the following actions after each grouping step: N-sized communities are removed from the graph, - for each node belonging to a community larger than N, the "loss of modularity" (as defined in said document) generated by the movement of the node to a community smaller than N is calculated, the motion of a node generating the lowest loss is applied, and the above process is reiterated as long as there are communities of size greater than N. After the segmentation of the graph into communities of size at the maximum N a power optimization algorithm is implemented, for example by maximizing a utility function for each community. This optimization can for example be implemented by the management entity or by an associated computing center. However, as a variant, it may be advantageous to distribute this calculation between the stations (nodes) of the same community, in order to increase the calculation speed while reducing the amount of signaling in the network. This second variant can be implemented as follows. During a step E5, the management entity informs each station of the structure of its community and of the distributed calculation that is incumbent upon it. More precisely: the joint distributions (mentioned above) are distributed within the same community, and the different initializations of the optimization calculation are distributed within the same community. During a step E6, each station performs an optimization calculation, and stops according to a predetermined convergence criterion.

Considering a community of size C, we denote by pki the level of the power offset ("offset" in English) of the node k with respect to a predetermined maximum power pmax for a central frequency subband. We denote pk the vector of pki components. Figure 3 illustrates these definitions in an example where C = 3. We define the following utility function: u (131, P2, -PC) = Eic = ift) Dk (il, j2, ic) ail ... pic , (2) Ekf # k ikf-Pkf where the function t () is a classical "link curve" giving the instantaneous descending rate for a user according to the signal-to-interference ratio. As defined above, Dk (ii, i2, ..., ic) is the joint distribution of interference perceived by the k-th station of the community and ik represents the useful power level of said station. Intuitively, the utility function represents the aggregated flows on the considered community (this quantity gives an importance equal to all the nodes, by not taking into account their duration of use). By discretizing the values of the vectors pk, it is possible to perform a combinatorial search for the optimum of the utility function u (p1, p2, ... pc). The number of values to be tested is however extremely high where q is the number of possible offsets and r the number of frequencies). It is therefore necessary to use a heuristic to solve this problem, for example to perform a genetic optimization. These methods have the good property of being easily parallelizable, for example using different initializations. It is of course possible to use alternative solutions based on greater cooperation between the nodes. During a step E7, each station disseminates the configuration of stations obtained following its calculation to the other members of its community if this configuration is more efficient than a previous or predetermined configuration. Finally, during a step E8, each station applies (for each sub-frequency band) the most efficient configuration among the set comprising the configuration obtained and the received configurations. Note finally that the use of an interference graph, as taught in the present invention, also to represent the influence of stations on the macro network. It is possible to integrate interference constraints on the macro network, 15 which facilitates the reuse of the frequencies of the latter, while maintaining its QoS level. As indicated above, the present invention also relates to a computer system implementing the station configuration method described above. This computer system 20 conventionally comprises a central processing unit controlling by signals a memory, as well as an input unit and an output unit. In addition, this computer system may be used to execute a computer program having instructions for implementing the method of configuring stations according to the invention. Indeed, the invention also relates to a downloadable computer program from a communication network comprising instructions for executing the steps of a station configuration method according to the invention, when it is executed on a computer. This computer program may be stored on a computer readable medium and may be executable by a microprocessor. This program can use any programming language, and present itself as source code, object code, or intermediate code between source code and object code, in a partially compiled form or in any other desirable form. The invention also relates to an information carrier, irremovable, or partially or completely removable, readable by a computer, and comprising instructions of a computer program as mentioned above. The information carrier may be any entity or device capable of storing the program. For example, the medium may comprise storage means, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or a magnetic recording medium, for example a USB flash drive ("USB flash drive"). in English) or a hard drive. On the other hand, the information medium may be a transmissible medium such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio or by other means. The computer program according to the invention can in particular be downloaded to an Internet type network. As a variant, the information carrier may be an integrated circuit in which the computer program is incorporated, the circuit being adapted to execute or to be used in the execution of the method according to the invention.

Claims (12)

REVENDICATIONS1. A method of configuring stations of a wireless communications network, said stations receiving according to a predetermined protocol the results of interference measurements from client devices to which they are connected, said method comprising the following step: a ) said stations transmit (E1, E2) data representative of said measurement results to a central management entity to which they are connected, said method being characterized in that it further comprises the following steps: b) said management entity aggregates (E3) said data to construct a global interference graph, c) the management entity segments (E4) the graph into communities of stations, d) performs (E5, E6, E7) an optimization calculation for each community, and e) each station applies (E8) the result of the optimization calculation which concerns it. 2. The method of configuring stations according to claim 1, characterized in that: during said step a), each of said stations aggregates (E1) said results received to obtain a joint distribution of interference in the form of a sub- the interference graph of which this station is the center, then sends (E2) its interference subgraph to said central management entity, and- during said step b), the management entity assembles (E3) the sub-graphs of interference interference graphs received to construct said global interference graph. 3. A method of configuring stations according to claim 1 or claim 2, characterized in that, in said step d): the management entity informs (E5) each station of the structure of its community and a distributed calculation that is incumbent on it, - each station performs (E6) an optimization calculation, and - each station disseminates (E7) the configuration of stations obtained following its calculation from the other members of its community if this configuration is more efficient than a preceding or predetermined configuration, and in that in said step e), each station (E8) applies the most efficient configuration of stations among the set comprising the configuration obtained and the configurations received. 4. Station configuration device of a wireless communications network, said station being able to receive, according to a predetermined protocol, the results of measurements of interference from client devices to which it is connected, said device being characterized in that it comprises means for: - aggregating said received results to obtain a joint interference distribution in the form of an interference sub-graph of which this station is the center, and - sending said sub-graph of interference to a central management entity to which it is linked. 5. station configuration device according to claim 4, characterized in that it further comprises means for: - receiving from a central management entity to which said station is connected information about the structure of a community of stations to which the station belongs and a distributed calculation which is incumbent upon it, - perform an optimization calculation, - disseminate the configuration of stations obtained following its calculation to the other members of said community if this configuration is more efficient than a previous or predetermined configuration, and - apply the most efficient configuration of stations among the set comprising the configuration obtained and the configurations received. Electronic circuit, characterized in that it comprises a station configuration device according to claim 4 or claim 5. 7. Central management entity in a wireless communications network, said management entity comprising means for receiving from the stations to which it is connected data representative of a joint interference distribution associated with each of these stations , characterized in that it further comprises means for: - aggregating said received data to construct a global interference graph, and - segmenting the graph into communities of stations. 8. Central management entity according to claim 7, characterized in that, said data received from each of said stations being in the form of an interference subgraph of which this station is the center, it comprises in in addition to means for assembling said interference sub-graphs to construct said overall interference graph. 9. Central management entity according to claim 7 or claim 8, characterized in that it further comprises means for informing each of said stations of the structure of its community and a distributed calculation which is incumbent on this station. 10. Electronic circuit, characterized in that it comprises a central management entity according to any one of claims 7 to 9. 11. Non-removable, or partially or completely removable data storage means comprising computer program code instructions for performing the steps of a station configuration method according to any one of claims 1 to 3. Computer program downloadable from a communication network and / or stored on a computer readable medium and / or executable by a microprocessor, characterized in that it comprises instructions for the execution of the steps of a method of station configuration according to any one of claims 1 to 3 when executed on a computer.