Method for determining a deployment configuration of access points for a wireless local area network in an area

Abstract

A method for determining a wireless local area network access point deployment configuration is disclosed, comprising: - obtaining, for access points that can be installed in an area, a first modeling error relating to a conversion between a performance criterion and a signal level, and a second modeling error relating to the signal level relative to signals emitted by the access points; - determining, for a pair of data routing paths passing through access points of a respective configuration, a probability that one path of said pair optimizes the performance criterion for said path relative to the performance criterion for another path of said pair, taking into account the modeling errors; - selecting a configuration for which an objective function of a performance criterion determined as a function of the probability is optimized. Abstract figure: Figure 1

Description

Title of the invention: Method for determining a deployment configuration of access points for a wireless local area network in an area technical field

[0001] The present disclosure falls within the domain of signal optimization in an area.

[0002] More specifically, this disclosure relates to a method for determining a wireless local area network deployment configuration in an area. Prior art

[0003] The locations and number of access points to a wireless local area network such as an Internet box and wifi repeaters in an area such as a home, company offices, etc. are important in order to ensure sufficient throughput at the receivers installed in the area.

[0004] A classic technique for placing access points in an area is experimental, and consists of placing or repositioning access points as best as possible based on the problems and coverage gaps encountered in the area under consideration. For this purpose, signal strength indicators or LED colors on the receivers of the signals sent by the access points can be used, and access points can be added to optimize a performance criterion such as the throughput at the receivers.

[0005] Another way of proceeding is to use a radio engineering tool (for example on a computer or with a mobile application) and optimization algorithms to define the minimum number of access points and their best locations, based on receiver requirements in terms of signal level and throughput.

[0006] Known solutions do not allow for taking into account uncertainties related to the different models used, for example, in engineering tools, and the deployment configuration determined for access points may therefore be suboptimal.

[0007] There is therefore a need for a method to overcome the drawbacks of the prior art and to obtain an optimal deployment configuration of access points (number and / or locations) to cover a given area. Summary

[0008] To this end, a method for determining a deployment configuration of wireless local area network access points in an area is disclosed, comprising: - obtaining, for access points likely to be installed in the area, a first modeling error (or prediction) relating to a conversion between values ​​of a network performance criterion and corresponding level values ​​of a signal that can be emitted by a respective access point, and obtaining a second modeling error (or prediction) relating to the signal level in relation to signals emitted by the access points in the area; - determination, for at least one pair of data routing paths among data routing paths passing through access points of a respective access point deployment configuration among a plurality of access point deployment configurations, of a probability that a data routing path of said at least one pair of data routing paths optimizes the performance criterion for said data routing path with respect to the performance criterion for another data routing path of said at least one pair of data routing paths, taking into account the first modeling error and the second modeling error; - selection of an access point deployment configuration for which an objective function of a performance criterion determined according to at least one probability is optimized.

[0009] An access point deployment configuration includes at least one primary access point (for example, a box or a gateway). Each access point deployment configuration can be defined by a number of access points and a placement of said access points.

[0010] Furthermore, one or more star and / or cascade repeaters behind the main access point may be installed in the area. The signal repeated by these repeaters may be a communication signal such as a Wi-Fi signal.

[0011] Different types of access points can be distinguished: - an Internet box connected directly to the Internet and capable of emitting a communication signal in the area; - an Ethernet repeater capable of receiving a communication signal provided by the Internet box or by another repeater in the area via a network socket to which the Ethernet repeater is connected by wired connection, and capable of transmitting a wireless communication signal to another repeater or receiver in the area; - a repeater called a "wireless repeater", connected wirelessly to the Internet box or to another repeater, and capable of transmitting a wireless signal to another repeater or a receiver in the area.

[0012] A data routing path can allow, in a given access point deployment configuration, the routing of a signal from a box Internet access is provided via a signal receiver installed in the area. Each data routing path can include multiple repeaters.

[0013] For a given pair of data routing paths, the data routing paths can carry a signal to the same receiver (signal destination) from the same access point (signal starting point), but via different repeaters.

[0014] Only one data routing path among the data routing paths considered may include only the Internet box.

[0015] For a given receiver, all combinations of data routing paths can be considered. For example, when three data routing paths cl, c2, c3 allow a signal to be routed to a given receiver from a given Internet box, pairs of data routing paths (cl, c2), (cl, c3), (c2, c3) can be considered, and probabilities can be determined for each of these pairs of data routing paths.

[0016] It is also possible to consider, for a given receiver, only a pair of data routing paths that includes the two data routing paths with the highest performance criterion. For example, when three data routing paths cl, c2, c3 allow a signal to be routed to a given receiver from a given Internet box, and when the performance criterion of data routing paths cl and c2 is greater than the performance criterion of data routing path c3, only the probability for the pair of data routing paths (cl, c2) can be determined.

[0017] A data routing path can be defined by the access points it includes.

[0018] Furthermore, a data routing path can be defined by the order of the access points. For example, the performance criterion of a data routing path where a signal passes, in order, through the access points AP2-AP3-AP1 may be different from the performance criterion of a data routing path where a signal passes, in order, through the access points AP2-AP1-AP3.

[0019] The performance criterion, also called a performance indicator, can be representative of the volume of information that a signal can carry.

[0020] The method can be implemented by a determination system.

[0021] Thus, "obtaining" can mean "receiving by the determination system." That is to say, the first modeling error and / or the second modeling error can be determined by an entity other than the determination system and be received by the determination system.

[0022] “Obtaining” can also mean “determining”. That is to say, the The first modeling error and / or the second modeling error can be determined by the determination system.

[0023] The second modeling error (or prediction) may be representative of the uncertainties and / or errors related to the modeling of the spatial distribution of the signal level in the area. The second modeling error may have the same value for the entire area, or different values, for example, an individual value for each access point or each link (a data routing path being composed of one or more links).

[0024] The first modeling error, also called the conversion error, can be representative of the uncertainties and / or errors related to the conversion of the signal level into the performance criterion. This first modeling error can be representative of the dispersion of the performance criterion values ​​for the same signal level value.

[0025] For example, by way of illustration, variability in temperature in the area (caused for example by an open window) or the movement of objects in the area (for example the opening of a door or the presence of a person) can affect said conversion and can be considered in the first modeling error.

[0026] The objective function of a performance criterion can be an indicator of the performance criterion related to the signals received by the receivers in the area. Thus, an access point deployment configuration that optimizes the objective function of a performance criterion can therefore be an access point deployment configuration in which the performance criterion of a signal is optimized at the receiver level.

[0027] The method is innovative in that it considers probabilities associated with the respective data routing paths as well as uncertainties of the models used, namely the first modeling error and the second modeling error.

[0028] The method makes it possible to select an access point deployment configuration that is optimal in terms of performance criteria for signals delivered to receivers and network range / coverage in the area. The performance criteria for signals received by the receivers can thus be maximized.

[0029] Access points are intended to be installed in the area according to the selected access point deployment configuration.

[0030] In one embodiment, the method may include: - Installation of access points in the area in accordance with the selected access point deployment configuration.

[0031] In one embodiment, each access point deployment configuration can be defined by a number of access points installed in the zone and / or by access points installed in the zone at defined locations.

[0032] Thus, the specific implementation of the process can be adapted to the circumstances and context of use.

[0033] In one example, when a fixed number of access points are intended to be placed in the area, the placement of the access points can be optimized.

[0034] In another example, a user may have a plurality of access points at his disposal, and determine the best deployment configuration (number and location) of the access points in the area.

[0035] In one embodiment, each access point deployment configuration can be defined by a number of access points installed at defined locations, and the selection of an access point deployment configuration can be performed in such a way as to optimize the objective function of a performance criterion while minimizing the number of access points.

[0036] Thus, the optimal deployment configuration of access points can be determined, i.e. the deployment configuration having a maximum performance criterion and a minimum complexity.

[0037] Furthermore, each access point deployment configuration can be characterized by the types of access points it includes.

[0038] In one embodiment, the process may include: - pre-selection of access point deployment configurations based on a constraint relating to said plurality of access points; - selection of the access point deployment configuration from the pre-selected access point deployment configurations.

[0039] For example, the constraint may relate to a power outlet to which the access point must be plugged. The access point can therefore be installed close to the power outlet.

[0040] The number of access point deployment configurations to consider in order to identify the optimal access point deployment configuration can thus be limited. The time required to determine the optimal access point deployment configuration can therefore be reduced.

[0041] In one embodiment, the selection of an access point deployment configuration is carried out at least in part by combinatorial optimization.

[0042] Thus, it is not necessary to determine the objective function of a performance criterion for each access point deployment configuration, but it is possible to discard unpromising access point deployment configurations and determine the objective function of a performance criterion only for access point deployment configurations deemed promising. The time required to determine the optimal access point deployment configuration can therefore be reduced.

[0043] In one embodiment, the method may include: - obtaining, for each access point, a ratio between the values ​​of a performance criterion and the power level values ​​of a signal that can be emitted by a respective access point; - determination of the objective function of a performance criterion taking into account said ratios.

[0044] When the signal level is simulated in the area (for example by using a propagation attenuation prediction model), knowledge of said ratio makes it easy and highly accurate to determine values ​​of the performance criterion from the signal level.

[0045] This way of determining the values ​​of a performance criterion makes it possible to create a model of a performance criterion that is significant and close to physical reality in the area.

[0046] The resulting objective function makes it possible to determine with high precision the optimal deployment configuration.

[0047] In one embodiment, the resulting report is in the form of a table comprising a plurality of signal level values ​​and corresponding performance criterion values ​​and / or in the form of an analytical function describing a relationship between signal level values ​​and corresponding performance criterion values.

[0048] In one embodiment, the analytical function is defined as: d(RSSI) = where RSSJ is a signal level value, d(RSSI) is a value of a corresponding performance criterion, eta, etc. are parameters depending on an access point type.

[0049] Thus, signal level values ​​can be determined for all or part of the area, and values ​​of a performance criterion can be determined from signal level values.

[0050] The advantage of using the table is that the value of a performance criterion can be taken directly for a signal level value included in the table, without additional calculation.

[0051] If a signal level value not contained in the table is sought, the corresponding performance criterion value can be obtained by interpolation with known neighbouring values.

[0052] The analytical function can be determined by a regression method applied to previously measured signal level values ​​and values ​​of a performance criterion for the signal of a respective access point. The parameters a, b, c can be determined by regression. Alternatively, other methods can be used to determine such an analytical function.

[0053] The advantage of the analytical function is that the value of a performance criterion can be determined directly from the function. Furthermore, the use of the analytical function allows for the analytical formulation of subsequent process operations.

[0054] In one embodiment, the objective function of a performance criterion is defined as: p ________1_______where N is a number of signal receivers installed in the fi+erf\2p-ï). zone, is the inverse error function, is a mean value deduced from a total performance criterion in the zone, is a standard deviation deduced from the total performance criterion in the zone, p is a confidence level value defined as p = Pr( 5 > S is the total performance criterion in the zone, and Pr( 5 > / >) is representative of a probability that S is greater than D.

[0055] An access point deployment configuration can be selected so as to guarantee a value of a performance criterion D of the signals routed to the receivers with a probability equal to or greater than p.

[0056] In one embodiment, the zone comprises a plurality of meshes, and is defined as: • \ L 1. 7 \ vn / y J j defined as: -, / 1 \ / 1 \ 2 where p / J. j is: e(4 = + ) e (74î?) ] NcMs is the number of meshes in the zone, ^sta(c^ 0 is the performance criterion of a signal routed by a data routing path cl from said at least pair of data routing paths to a mesh h and dSTA(c2,1) is the performance criterion of a signal routed by another data routing path c2 from said at least pair of data routing paths to the mesh », where is a defined as: ç--1—, where £> / is, for a respective receiver j among the signal receivers, a value of a performance criterion routed to said receiver j when said receiver j is the only receiver in the area, jyi being a function of a receiver location j, and P^i) is a probability that a value of a performance criterion routed by a data routing path cl among said at least one pair of data routing paths to a mesh i is greater than a value of a performance criterion routed by another data routing path c2 of said at least one pair of data routing paths to said mesh.

[0057] The number of meshes and the nature of the meshes for a given area can be left to the user's choice according to the environment considered and the desired resolution.

[0058] In one embodiment, P y ( i ) is defined as: Pi(i) = Pr(L / i)~ L^ + G / i) <Cy ^Pr^L^- Ui)+ G^i) <C) est représentative d’une probabilité que L [i) - Lÿ) + G^) is less than C, TV”1 Z1 ' Li[i]= G^g^g- _Vh2J_ , “ ^n'^n' c^. —q RSSI +b - ln(c)' C”' are the parameters of the function analytical, j —____________^sï~T is a signal level value relating to a link n', g^g? are centered Gaussian random variables with standard deviations that are a function of the respective first modeling error, wl, ni are a number of links comprising a respective data routing path among said at least one pair of data routing paths.

[0059] p (---1— \ and p (___L— \ correspond to the expectation of---1— and ---1— respectively. ustaW>

[0060] Each data routing path can include one or more links. A link can be a direct connection between two access points, i.e., a connection that includes these two access points but does not pass through any other access points.

[0061] The objective function of a performance criterion can take into account the first modeling error and the second modeling error via the probability P y ( i ) as well as via p / -----1-----j and p / ______î-----j .

[0062] Another aspect of this disclosure relates to a computer program product comprising instructions which, when executed by a processor, cause the processor to carry out operations of the aforementioned process.

[0063] This program may use any programming language (for example, an object-oriented language or other), and may be in the form of interpretable source code, partially compiled code or fully compiled code.

[0064] Another aspect of the disclosure relates to a system for determining a deployment configuration of wireless local area network access points in an area, comprising: - an interface configured to obtain, for access points that may be installed in the area, a first modeling error relating to a conversion between values ​​of a network performance criterion and corresponding level values ​​of a signal that may be emitted by a respective access point, and to obtain a second modeling error relating to the signal level relative to signals emitted by the access points in the area; - a circuit configured to determine, for at least one pair of data routing paths among data routing paths passing through access points of a respective access point deployment configuration among a plurality of access point deployment configurations, a probability that a data routing path of said at least one pair of data routing paths optimizes the performance criterion for said data routing path with respect to the performance criterion for another data routing path of said at least one pair of data routing paths, taking into account the first modeling error and the second modeling error; - a circuit configured to select an access point deployment configuration for which an objective function of a performance criterion determined based on at least one probability is optimized.

[0065] The system can be configured to implement the aforementioned process.

[0066] The system is innovative in that it considers probabilities associated with the respective data routing paths as well as uncertainties of the models used, namely the first modeling error and the second modeling error.

[0067] The system allows the selection of an access point deployment configuration that is optimal in terms of performance criteria for signals delivered to receivers and network range / coverage in the area. The performance criteria for signals received by the receivers can thus be maximized. Brief description of the drawings

[0068] Other features, details and advantages will become apparent from the detailed description below and from the analysis of the accompanying drawings, in which: Fig. 1

[0069] [Fig-1] is a schematic representation of an area for which an access point deployment configuration is determined by a method for determining an access point deployment configuration of a local area network in an area. Fig. 2

[0070] [Fig.2] shows the flowchart of the process. Fig. 3

[0071] [Fig.3] is a schematic representation of a relationship between signal level values ​​and values ​​of a performance criterion of a signal that can be emitted by a given access point. Fig. 4

[0072] [Fig.4] is a schematic representation of an embodiment of a system for determining a deployment configuration of access points of a local network in an area configured to implement the method. Description of the implementation methods

[0073] The present disclosure relates to a method and system for determining a deployment configuration of wireless local area network access points in an area.

[0074] The determination of the access point deployment configuration is carried out on the basis of an objective function of a performance criterion taking into account a first modeling error (or prediction) and a second modeling error (or prediction).

[0075] Figure 1 shows an area Z, such as a home or office, for which the method can be implemented. Of course, these examples are given only for illustrative purposes, and the invention can be applied to any area intended to be covered by a wireless local area network.

[0076] In the example envisaged in [Fig.1], REC1, REC2 receivers of a signal (e.g. of a wifi signal) such as a computer or a mobile phone are installed in zone Z at defined locations, and a deployment configuration of API access points, AP2, AP3 such as an Internet box or repeaters is sought to maximize the signal level at the level of the REC1, REC2 receivers.

[0077] An access point deployment configuration can be defined by a number of access points and a placement of said access points.

[0078] The API access point, here in the form of an Internet box, can route a signal directly to the REC1 receiver via a direct data routing path cl or indirectly via a data routing path c3 through AP2 or a data routing path c2 through AP2.

[0079] For clarity, the data routing paths to the REC2 receiver are not shown.

[0080] [Fig.2] shows a flowchart of process 100. Process 100 can be implemented by a SYS system as shown in [Fig.4].

[0081] In a first operation 101 of the process 100, the SYS system can obtain a surface plane P of the zone Z. For example, the surface plane P can be downloaded by the SYS system or detected by the SYS system.

[0082] The surface plane P of the area Z may comprise a plurality of cells C. These cells C may have different sizes and / or shapes. The cells C may be pixels. For example, the size of a cell C may be between 50 cm and 100 cm.

[0083] Furthermore, the SYS system can obtain information relating to a deployment configuration of the REC1, REC2 receivers installed in the Z zone, for example location coordinates of the REC1, REC2 receivers in the Z zone.

[0084] In addition, the SYS system can obtain information characterizing the signal that can be emitted by the API, AP2, AP3 access points that may be installed in the Z zone. In particular, the SYS system can obtain a relationship between signal level values ​​and values ​​of a performance criterion of a signal that can be emitted by a respective API, AP2, AP3 access point.

[0085] Figure 3 shows an example of the ratio R in the form of an analytical function allowing the determination, for a given signal, of a value of a performance criterion as a function of a signal level value or vice versa.

[0086] The signal level may have as its unit "dBm", or be an intensity (typically indicated in W / cm2) or a relative intensity representative of a signal strength in the Z zone as a function of location.

[0087] The values ​​of a performance criterion may correspond to throughput values ​​and have as their unit "bps" (bits per second).

[0088] Alternatively, the values ​​of a performance criterion may, for example, correspond to values ​​relating to the latency of signal transmission, or to the quality of the signal when the signal is audio or video content.

[0089] The analytical function can be determined by a regression method applied to signal level values ​​and values ​​of a performance criterion previously measured for the signal of a respective API, AP2, AP3 access point. For example, the signal level can be measured with a spectrum analyzer.

[0090] The analytical function can be in the following form: d(RSSI) (eq. 1) where RSSI (for "Received Signal Strength Indicator") are signal level values, d(RSSI) are performance criterion values, and a, b, c are parameters determined by regression.

[0091] In an alternative not shown in the figures, the R ratio may be in tabular form comprising signal level values ​​and corresponding performance criterion values.

[0092] If a signal level value not contained in the table is sought, the corresponding performance criterion value can be obtained by interpolation with known neighbouring values.

[0093] Returning to the first step 101 of the process in [Fig. 2], the SYS system can also obtain a first modeling error related to a conversion between values ​​of a performance criterion and corresponding level values ​​of a signal that can be emitted by a respective API, AP2, or AP3 access point. The first modeling error can be representative of a dispersion in the values ​​of a performance criterion.

[0094] Furthermore, a second modeling error can be obtained. The second modeling error can be representative of the uncertainties and / or errors related to the modeling of the spatial distribution of the signal level in the Z zone. The second modeling error can have the same value for the entire Z zone, or different values, for example, an individual value related to each access point API, AP2, AP3.

[0095] Depending on the model used, the first modeling error / second modeling error may be standard deviations.

[0096] Different access point deployment configurations in the Z zone are considered, and the objective function of a performance criterion (which will be discussed later in more detail) of the signals at the REC1, REC2 receivers can be determined for each access point deployment configuration.

[0097] Each access point deployment configuration can be characterized by the number of API, AP2, AP3 access points it includes and by the locations at which the API, AP2, AP3 access points are installed.

[0098] In a second operation 102 of the process 100, a propagation of the signal can be simulated in the Z zone in order to determine a signal level in the Z zone, i.e. a spatial distribution of the signal level, using a propagation attenuation prediction model.

[0099] Such a simulation can be done for a plurality of access point deployment configurations.

[0100] In the simulation, the transmission of the signal from one access point API, AP2, AP3 to the other and the routing of the signal to the receivers REC1, REC2 is considered for a respective access point deployment configuration.

[0101] As explained, the second modeling error can be obtained in the first step 101 of the process 100. Alternatively, the second modeling error can be determined in the second step 102 during the simulation of the signal level in the Z zone. In this case, the second modeling error can take into account specific circumstances in the Z zone such as the presence of walls.

[0102] Different data routing paths cl, c2, c3 may exist to route a signal from an Internet API box to a REC1, REC2 receiver. That is to say, a signal emitted by an API access point may pass through different access points AP2, AP3 before arriving at the REC1, REC2 receivers.

[0103] When a signal passes successively through several API access points, AP2, AP3, the performance criterion of a respective data routing path cl, c2, c3 can be determined as the harmonic mean of the performance criteria of the individual links 112, lu, I23, 1ir, I2R, I3R that make up a respective data routing path cl, c2, c3: where dSTA is the performance criterion of a data routing path cl, c2, c3 composed of several access points API, AP2, AP3, and dn is the performance criterion of a signal transmitted by a link 112, li3, 123, liR, 12R, 13R respectively and Niink is the number of links In, lis, I23, 1ir, I2R, I3R through which the signal passes.

[0104] A link b2, b3, 123, 11R, 12R, 13R can be a direct connection between two API access points, AP2, AP3, i.e. a connection which involves the two respective API access points, AP2, AP3 but which does not pass through other API access points, AP2, AP3.

[0105] In [Fig. 1], a signal can be routed from an API access point (e.g., an Internet box) to a REC1 receiver either via a data routing path c1 (links 1n and 13R) via an AP3 access point (e.g., a repeater) or via another data routing path c2 via an AP2 access point (e.g., a repeater), i.e., via links b2 and 12R. Alternatively, the signal can be transmitted directly to the REC1 receiver via link 1[R.

[0106] The performance criterion for the data routing path cl is therefore - ......1...... and the performance criterion of the other data routing path c2 is

[0107] In a third operation 103 of the process 100 according to [Fig.2], for each access point deployment configuration, a result of an objective function of a performance criterion can be determined, taking into account the first modeling error and the second modeling error.

[0108] The access point deployment configuration for which the objective function of a performance criterion is optimized can be selected.

[0109] In what follows, an expression for the objective function of a performance criterion which takes into account the first modeling error and the second modeling error is deduced.

[0110] A total performance criterion S for zone Z can be defined as: S =---i—: (equ. 3) where N is a number of REC1, REC2 receivers in the Z zone. is, for a respective receiver j among the REC1, REC2 signal receivers, a value of a performance criterion routed to said receiver j when said receiver j is the only receiver in the Z zone, jÿ being a function of a location of receiver j.

[0111] / y is a random variable because the locations of the REC1, REC2 receivers can be random.

[0112] By applying the central limit theorem, the inverse 1 / S of S can be modeled according to the normal distribution, with a mean p and a variance N. <r, sur la base de l’écart type o des variables aléatoires hjÿ distribuées manière uniforme les mailles c au nombre ncens dans zone z, où p et sont définis comme : 1 y'Np^1 1 , 1 1 \, (equ. 4) ~N pixel D; ~Npixel \Dt ~ P]

[0113] A' corresponds to the best performance criterion available at the level of a mesh i among the performance criteria of signals routed via all the data routing paths cl, c2, c3 allowing a signal to be routed to said mesh i.

[0114] As explained, a given REC1, REC2 receiver can receive a signal via different data routing paths cl, c2, c3.

[0115] Pairs of data routing paths (cl, c2), (cl, c3), (c2, c3) can be considered, where each data routing path cl, c2, c3 of a pair of data routing paths starts from an Internet API box and ends on the same REC1 receiver.

[0116] By way of example, it is considered that the performance criterion of the data routing paths cl, c2 is superior to the performance criterion of the data routing path c3, and only the pair with the data routing paths cl, c2 is therefore considered.

[0117] The probability PI is determined that a data routing path cl of the pair of data routing paths (cl, c2) optimizes a performance criterion with respect to the performance criterion of the other data routing path c2 of the pair of data routing paths (cl, c2). Pl = Pr(dSTA(cl) = '

[0118] PI is a function of the mesh i to which a data routing path cl, c2, c3 leads. For clarity, the parameter i is omitted in the calculations below.

[0119] When the ratio R is expressed as an analytic function according to equation 1, PI can be expressed as: » ^Nlink.2 l+e^a^SSI^+b;) \ ) (equ. 5) Pl = PrL. -----, -----T5----- + g2 where ai bl, ci ai cl are parameters of the analytical function determined by regressions, and g^ and g^ are Gaussian random variables with a mean equal to 0 and an error as a function of the first respective modeling error.

[0120] RSSIft and RSSI„ are Gaussian variables with a standard deviation equal to the first modeling error.

[0121] In the case where the data routing paths cl, c2 have links in common, the common links can be removed and the API access points, AP2, AP3 of the two data routing paths cl, c2 can be reindexed.

[0122] After removing the common links, PI becomes: m tJv"1 I+expfeRS^ । v"2 ^^IJlSSI^+b^ \ (equ. 6) Pl = Pn2_ , .------,-----*+g, <l , ,-----7------+g2,

[0123] “1, “2 are defined as the number of links composing a routing path of data cl, c2 respectively.

[0124] By defining yj = an.KSSl„.+b„, = °U -J » 2 2 \ represents a Gaussian random variable with a mean equal to aJf*Rssi ict a variance equal to n-2 , where corresponds to the 'n' RSSIji' second modeling error of a link n', PI deviates: h! । " If n! ci-

[0125]

[0126] The links n of a given data routing path can have the same value aRSShi\ By defining: fl fl H fl II fl \ «^RSSI / i + hV «7' an ^RSSIji \ Yv;iir^n2 ।

[0127] Z. PI deviant : Pl = Pr(Lr L2 + G3 <c) (équ. 7) _ V"1 iy"2 2 is a Gaussian random variable with a mean equal to 0 and a standard deviation defined as: ( n[ + n2 - 2 ) ct ^6,2 are The initial modeling errors relate to the two types of links present (links between two access points and links between an access point and a receiver; MIMO being able to

[0128]

[0129]

[0130]

[0131] (be different, 4x4 for example for the 1st case and 2x2 for a typical receptor). g^ g? are independent Gaussian variables with a mean equal to 0 and a standard deviation equal to the first modeling error of the respective link. L1 and L2 can be approximated by logarithmic normal distributions using the Fenton-Wilkinson approximation method. Monte Carlo simulations based on logarithmic normal distributions can be applied to determine PI based on equation 7. In the example above, it was assumed that the ratio R is an analytical function according to equation 1. When the R ratio is in tabular form, in order to determine the values -L of equation 5, Monte Carlo simulations are implemented: values ​​a» Gaussian signal level values ​​can be randomly selected; the values ​​-L are determined from the table, and the first modeling errors obtained from Gaussian samples with a mean and standard deviation corresponding to the first modeling errors are added. The objective of method 100 is to determine an access point deployment configuration for which the total performance criterion S of signals routed to the N receivers REC1, REC2 in zone Z is greater than a criterion of performance Æ with a certain probability, i.e. with a confidence level p: 2 (eq. 8) = P D=________1 (eq. 9) fi+erf \lp-1).

[0132] Equation 9 corresponds to the objective function of a performance criterion mentioned above, assuming that the repeaters are all wireless repeaters (such as wifi repeaters, for example in the case of a WiFi-mesh network sharing the same channel).

[0133] For example, a value of p = 95% may mean that a total performance criterion S greater than a value D is routed to the N receivers REC1, REC2 in the Z area with a probability of p = 95%.

[0134] The different variables and parameters of the objective function of a performance criterion are defined as follows: = E(i)= [ P1 ( * ) E dsTA(<^ ) + ( ( O ) E dSTA{c2,ï}2 ) ] ^9 _ p 1 jp J. 2 °where and are respectively a mean and a standard deviation deduced from the performance criterion S and where p / ---1— ) is defined as: e(7-m)=eUU, ---3---+ ^)2)

[0135] Similar definitions apply to p / ---lj and p / ----1__\.

[0136] With regard to the expression p / 1___) , the two cases where the paths of \ 'East+G-'L) / Data routing cl, c2 have links in common (n=n') and where they have no links in common (n^n) can be considered.

[0137] When n^n': b [-------ç-----j = 44 + ïî ^a'„.RSS / n + b„ + —5— I+ 4 exp|a^.KSS / n. +bn.+ —5— ' t ^«fksss / i+4++^jess^+4+=e"'

[0138] When n=n': I ( l-texp(<4«5Sr„+»0 \ \ 1 I i ( ;d <w ti °iss tâ j (^jnccri 4¾^)2 \ 1 F \ (-----4-----) + 2exP v +bn+—2— ] + exp [ 2anJtSSIn+2bn + —j— ) ] =Cn,n

[0139] Thus, we obtain for g / ----J._____- ) : V d^ / palh^ 1 / ) / / i\j-'NIinkj w-iNlinkir-iNlinki E[ ——ï—ttâ)=L । €„# + 2^ . XL- ,en„'+(NlinkJ-1)(7^, +(7^ d2STA\path_U) / ^n=ï rv' ^n-n+ï riji \ ?

[0140] To return to process 100 according to Figure 2, in the third operation 103 of process 100, for each access point deployment configuration among a plurality of access point deployment configurations, a result of an objective function of a performance criterion can be determined, taking into account the second modeling errors (7^$$^ and the first modeling errors

[0141] The access point deployment configuration that maximizes the value D of the objective function of a performance criterion, called the optimal access point deployment configuration, can be selected.

[0142] The optimal access point deployment configuration can be determined by combinatorial optimization, by determining the value D of the objective function of a performance criterion for access point deployment configurations considered promising.

[0143] The number of access point deployment configurations to be tested, i.e., the number of access point deployment configurations for which the objective function of a performance criterion is determined, can be reduced based on one or more constraints such as power outlets. Indeed, an API, AP2, AP3 access point is typically plugged into a power outlet and will normally be located close to the outlet, which considerably limits the number of possible locations for this API, AP2, AP3 access point.

[0144] Certain access point deployment configurations can therefore be preselected based on the constraint.

[0145] In a fourth operation 104, API access points, AP2, AP3 can be installed in the Z zone in accordance with the selected access point deployment configuration.

[0146] Fig. 4 shows a SYS system mode configured to implement at least part of the determination process 100.

[0147] The SYS system includes at least one input interface 201 for receiving messages or instructions, such as the surface plane P and the ratio R for the API access points, AP2, AP3, and at least one output interface 202 for communication with external devices 205.

[0148] The SYS system further includes a memory 203 for storing instructions enabling the implementation of at least part of the process 100, the received data, and temporary data for carrying out the various operations 101, 102, 103, 104 of the process 100 as described above.

[0149] The SYS system further includes a circuit 204. This circuit 204 may be, for example: - a processor capable of interpreting instructions in the form of a computer program, or - an electronic board whose operations of process 100 of disclosure can be described in silicon, or even - a programmable electronic chip such as an FPGA chip for "Field-Programmable Gate Array" in English, such as a SoC for "System On Chip" in English or such as an ASIC for "Application Specified Integrated Circuit" in English.

[0150] The same circuit 204 can be used to put the operations 101, 102, 103, 104 of the process 100 or of individual circuits 204.

[0151] Depending on the embodiment, the SYS system may be a computer, a computer network, an electronic component, or another device comprising a processor operationally coupled to memory, and, depending on the chosen embodiment, a data storage unit, and other associated hardware elements such as a network interface and a media reader for reading and writing to removable storage media not shown in [Fig. 4]. The removable storage media may be, for example, a compact disc (CD), a digital video / multipurpose disc (DVD), a flash drive, a USB flash drive, etc.

[0152] In particular, the SYS system can be a computer or a smartphone. Thus, a user can easily optimize the deployment configuration of their Internet box and repeaters in their home by executing an algorithm that implements operations 101, 102, 103, 104 of process 100 on their computer or smartphone.

[0153] Depending on the embodiment, the memory 203, the data storage unit or the removable storage medium contains instructions which, when executed by the circuit 204, cause this circuit 204 to perform or control at least one input interface 201, at least one output interface 202, the storage of data in the memory 203 and / or the processing of data and / or the implementation of at least one part of the method 100 according to [Fig.2].

[0154] Circuit 204 can be a component implementing the control of the SYS system.

[0155] Furthermore, the SYS system can be implemented in software form, in which case it takes the form of a program executable by a processor, or in hardware form, such as an application-specific integrated circuit (ASIC), a system-on-chip (SOC), or as a combination of hardware and software, for example a software program intended to be loaded and executed on an electronic component described above such as an FPGA.

[0156] The SYS system can also use hybrid architectures, for example architectures based on a CPU+FPGA, a GPU for “Graphics Processing Unit” or an MPPA for “Multi-Purpose Processor Array”.

[0157] This disclosure is not limited to the examples of devices, systems, processes, uses and computer program products described above, only by way of example, but encompasses all the variations that a person skilled in the art may consider in the context of the protection sought.< / w> < / l>

Claims

Demands

1. Method (100) for determining a deployment configuration of access points (API, AP2, AP3) of a wireless local area network in an area (Z), comprising: - obtaining, for access points (API, AP2, AP3) that can be installed in the area (Z), a first modeling error relating to a conversion between values ​​of a network performance criterion and corresponding level values ​​of a signal that can be emitted by a respective access point (API, AP2, AP3), and obtaining a second modeling error relating to the signal level with respect to signals emitted by the access points (API, AP2, AP3) in the area (Z);- determination, for at least one pair of data routing paths (cl, c2) among data routing paths (cl, c2, c3) passing through access points (API, AP2, AP3) of a respective access point deployment configuration among a plurality of access point deployment configurations, of a probability that a data routing path (cl) of said at least one pair of data routing paths (cl, c2) optimizes the performance criterion for said data routing path (c2) with respect to the performance criterion for another data routing path (c2) of said at least one pair of data routing paths (cl, c2), taking into account the first modeling error and the second modeling error; - selection of an access point deployment configuration for which an objective function of a performance criterion determined as a function of the at least one probability is optimized.

2. Method according to claim 1, comprising: - installation (104) of access points (API, AP2, AP3) in the zone (Z) in accordance with the selected access point deployment configuration.

3. A method according to any one of the preceding claims, wherein each access point deployment configuration is defined by a number of access points (API, AP2, AP3) installed in the zone (Z) and / or by access points (API, AP2, AP3) installed in the zone (Z) at defined locations.

4. A method according to any one of the preceding claims, comprising: - pre-selecting access point deployment configurations based on a constraint relating to said plurality of access points (API, AP2, AP3); - selecting the access point deployment configuration from among the pre-selected access point deployment configurations.

5. A method according to any one of the preceding claims, wherein the selection of an access point deployment configuration is carried out at least in part by combinatorial optimization.

6. A method according to any one of the preceding claims, comprising: - obtaining, for each access point (API, AP2, AP3), a ratio (R) between values ​​of a performance criterion and power level values ​​of a signal that can be emitted by a respective access point (API, AP2, AP3); - determining the objective function of a performance criterion taking into account said ratios (R).

7. A method according to any one of the preceding claims, wherein the objective function of a performance criterion is defined as: 2) —_______J_______ where is a number of receivers (REC1, P+erf^p-ï). REC2) of signals installed in the area (Z), erf1 is the inverse error function, is a mean value deduced from a total performance criterion in the area (Z), n is a standard deviation deduced from the total performance criterion in the area (Z), p is a confidence level value defined as p = Pr( S > D), where S is the total performance criterion in the area (Z), and Pr( S > D) represents a probability that S is greater than D.

8. A method according to claim 7, wherein the zone (Z) comprises a plurality of meshes (C), and wherein p is defined as: / <= E(i) = is defined as: , ■ / 1 \ , ■ / [ \2 where p / J- j is: <72= E^ ) - E(E'V[P}(i)E() + (1 - Pi(i))E(...)] where NceUs is the number of meshes (C) in the area (Z), dSTA(c1, i) is the performance criterion of a signal routed by a data routing path c1 from said at least pair of data routing paths (c1, c2) to a mesh*, and dSTA(c2, i) is the performance criterion of a signal routed by another data routing path c2 from said at least pair of data routing paths (c1, c2) to mesh i, where is a defined as: ç--1---, where pi is, for a respective receiver j among the signal receivers (REC1, REC2), a value of a performance criterion routed to said receiver j when said receiver j is the only receiver in the area (Z), / y being a function of a location of receiver j, and P^i} is a probability that a value of a performance criterion is routed through a data routing path (cl,(c2) among said at least one pair of data routing paths (cl, c2) to mesh i is greater than a value of a performance criterion routed by another data routing path (cl, c2) of said at least one pair of data routing paths (cl, c2) to said mesh i.,

9. Product computer program comprising instructions which, when executed by a processor, cause the processor to carry out operations of a process (100) according to any one of claims 1 to 8.

10. System (SYS) for determining a deployment configuration of access points of a wireless local area network in an area (Z), comprising: - an interface configured to obtain, for access points (API, AP2, AP3) that may be installed in the area (Z), a first modeling error relating to a conversion between values ​​of a network performance criterion and corresponding level values ​​of a signal that may be emitted by a respective access point (API, AP2, AP3), and to obtain a second modeling error relating to the signal level with respect to signals emitted by the access points (API, AP2, AP3) in the area (Z); - a circuit (204) configured to determine, for at least one pair of data routing paths (cl, c2) among data routing paths (cl, c2, c3) passing through access points (API, AP2, AP3) of a respective access point deployment configuration among a plurality of access point deployment configurations, a probability that a data routing path (cl) of said at least one pair of data routing paths (cl, c2) optimizes the performance criterion for said data routing path (c2) with respect to the performance criterion for another data routing path (c2) of said at least one pair of data routing paths (cl, c2), taking into account the first modeling error and the second modeling error; - a circuit (204) configured to select an access point deployment configuration for which an objective function of a performance criterion determined according to at least one probability is optimized.

Images