Method and server for configuring parameters of formation moving in vicinity

Abstract

The present disclosure relates to a computer-implemented method for configuring values of a set of parameters of at least two mobile formations, each formation comprising a plurality of respective communication nodes, where the set of parameters comprises at least one priori parameter, where, for each formation, the priori parameters are associated with each other. The set of parameters further comprising: control parameters of a control algorithm for configuring communication parameters of the wireless link, the configuration method comprising the steps of:-for each priori parameter, obtaining a range of allowable performance values based on a corresponding priori parameter value; searching for values of control parameters and values of communication parameters of the formation capable of achieving allowable performance values for each range in an optimal manner; -if it is impossible to achieve the permissible performance value for each range, searching for an updated range of permissible performance values for which a value of the control parameter and a value of the communication parameter capable of achieving the permissible performance value for each updated range can be determined; -determining an updated value of the at least one priori parameter.

Description

Technical Field

[0001] The present disclosure relates to communication systems, and more particularly to methods and servers for configuring parameter sets of at least two formations of communication nodes moving toward each other in proximity. Background Art

[0002] Many applications can benefit from building a platoon of multiple vehicles that need to follow the same overall trajectory while cooperatively maintaining a desired geometric arrangement among themselves. Moving in a platoon has many advantages over non-cooperative systems to reduce system cost, increase system robustness and efficiency, while providing redundancy, reconfiguration capabilities, and architectural flexibility [Chen2005].

[0003] Among them, platooning appears as a very promising application, where several vehicles are required to travel at a desired speed while maintaining a specified distance between the vehicles (e.g., 1D on a lane). Driving in a platoon of cars, trucks, robots, or even airplanes may lead to significant benefits in terms of power consumption, passenger comfort, traffic smoothness, etc. Other kinds of platooning applications may be satellite clusters, drone swarms for security, search and rescue, agriculture, etc.

[0004] In order to maintain the geometric arrangement of the formation, the dynamics of the vehicles in the formation are usually handled using control algorithms that require the exchange of data between communication nodes embedded in the vehicles. For example, a platoon usually implies the transmission of the acceleration and velocity of the lead vehicle to its followers and the position and velocity of each vehicle to its neighbors (see, e.g., [Sybis2019]). Then, for example, a consensus algorithm can be used to keep the vehicles at equal distances from each other.

[0005] Most platooning applications rely on wireless links due to their mobility requirements. Communications over wireless links are prone to errors from physical layer issues (additive noise, path loss, shadowing, fast fading, phase noise, etc.), but also from upper layer issues (collisions due to hidden node effects in carrier sensing multiple access (CSMA) schemes or distributed scheduling, etc.). Delays are also caused to some extent by physical layer effects (transmission delays, packet duration, etc.), but mostly by upper layer effects (scheduling strategies, random access schemes, etc.).

[0006] By the control algorithms depending on the data exchange between the communication nodes, the loss of data packets is obviously detrimental to their performance. Packets may also be lost due to a shortage of communication resources. Due to the scarcity of radio frequencies, wireless systems are designed to operate on a limited bandwidth. Regardless of the transmission technology used, this results in only a limited number of communication resources being available to the communication nodes per time unit (either at the physical level or at the logical level where the operator dedicates only a subset of the overall available resources to a specific application).

[0007] When a single platoon of vehicles is moving while maintaining a predetermined geometric arrangement between the vehicles (typically 1D with a predetermined distance and predetermined speed between them) using a control algorithm (such as a consensus algorithm) and limited communication resources, it is important to ensure that the control algorithm can actually control the shape of the geometric arrangement and / or its speed with a sufficient convergence speed. The convergence speed of the control algorithm depends, among other things, on the rate of packet loss (see, e.g., [Sybis2019]).

[0008] However, the formation needs to be able to anticipate modifications of the communication environment, which may prevent achieving sufficient convergence speed given the geometric arrangement of the formation.

[0009] Such modification of the communication environment may occur, for example, when different formations are in proximity to each other and use the same communication resources to exchange data related to their respective control algorithms. As the formations move into each other's radio coverage, the communication nodes of the formations will compete for access to the same communication resources, resulting in increased mutual interference and increased packet loss. Examples of formations that may move into radio proximity and compete for the same communication resources include convoys of vehicles crossing or overtaking each other on a highway. Summary of the invention

[0010] The present disclosure aims to improve the situation. In particular, the present disclosure aims to overcome at least some limitations of the above-mentioned prior art by proposing a solution of a set of parameters for configuring at least two formations moving in proximity to each other.

[0011] To this end, and according to a first aspect, the present disclosure relates to a computer-implemented method for configuring values ​​of a set of parameters for at least two mobile formations, each formation comprising a plurality of corresponding communication nodes, wherein the set of parameters comprises at least one prior parameter, wherein each prior parameter corresponds to a formation parameter of each formation or a network parameter of the formation, wherein the formation parameters of the formation define a geometric arrangement of the communication nodes of the formation and the network parameters of the formation define communication resources allocated to the formation for exchanging data over a wireless link, wherein each prior parameter has a predetermined value and a predetermined change in the communication environment of the formation is about to occur. For each formation, the set of parameters further comprises:

[0012] - control parameters of the control algorithm used by the formation to maintain the geometric arrangement,

[0013] - communication parameters for configuring the radio links between the communication nodes of the formation, said radio links being used to exchange data related to the control algorithm,

[0014] The configuration method comprises the following steps:

[0015] - for each a priori parameter, obtaining a range of permissible performance values ​​based on the corresponding a priori parameter value,

[0016] - the values ​​of control parameters and communication parameters of the search formation capable of achieving permissible performance values ​​in each range, taking into account the modifications that have occurred in the communication environment,

[0017] - if it is not possible to achieve the permissible performance value for each range, searching for an updated range of permissible performance values ​​for which values ​​of control parameters and values ​​of communication parameters can be determined that enable the permissible performance value for each updated range to be achieved, taking into account that a change in the communication environment has occurred,

[0018] - determining an updated value of at least one a priori parameter based on the value of the control parameter and the value of the communication parameter, said updated value enabling the achievement of an allowed performance value for each updated range of allowed performance values.

[0019] Therefore, the proposed configuration method first considers at least one a priori parameter as an input. The a priori parameter corresponds to a parameter having a value configured before a predetermined change in the communication environment of the formation is about to occur. For example, the expected change in the communication environment is due to the fact that the formations are approaching each other and will be near the radio and may compete for the same communication resources allocated. According to another example, the expected change in the communication environment is due to the fact that, for example, when moving near the radio with at least one other formation, at least one formation will experience a predictable degradation of the quality of its wireless link. In the present disclosure, when a change in the communication environment occurs, the value of each a priori parameter is maintained (if possible). For example, a priori parameters corresponding to the formation parameters of each formation can be considered, which describe the geometric arrangement of the formation (relative position between communication nodes, speed of communication nodes, etc.). In this case, the value of the formation parameter (geometric arrangement) should remain unchanged (if possible). In some cases, a priori parameters corresponding to network parameters describing the amount of communication resources allocated to one or more formations can also be considered. In this case, the value of the network parameter (the amount of allocated communication resources) should remain unchanged (if possible).

[0020] The range of allowed performance values ​​is determined based on the value of the a priori parameter. The allowed performance value is essentially the value of the performance flag that enables the value of the a priori parameter to be maintained. Before a modification of the communication environment of the formation occurs (e.g., before the formations are in the vicinity of each other), the performance flags have values ​​within their respective allowed ranges.

[0021] Then, before a change in the communication environment occurs (or shortly after its onset), but taking into account that said change in the communication environment has occurred, the configuration method searches for values ​​of communication parameters (for configuring the radio links) and values ​​of control parameters (for configuring the control algorithms of the formation) that enable the achievement of an allowable performance value for each range considered. The search can be performed, for example, by any multi-objective optimization method known to those skilled in the art.

[0022] If such values ​​of the communication parameters and the control parameters exist, this means that the expected change of the communication environment does not prevent maintaining the values ​​of the a priori parameters (eg maintaining the geometric configuration).

[0023] If such values ​​of the communication parameters and the control parameters do not exist, this means that an expected change in the communication environment prevents maintaining the values ​​of the a priori parameters (eg maintaining the geometric configuration).

[0024] In the latter case, the configuration method searches for an updated range of allowed performance values ​​for which it can be determined that the values ​​of the control parameters and the values ​​of the communication parameters of the allowed performance values ​​can be achieved for each updated range, taking into account the changes in the communication environment that have occurred. Therefore, the ranges are iteratively updated by modifying (i.e., downgrading) at least one range until an achievable allowed performance value is obtained for each updated range. Modifying the range of allowed performance values ​​means including in the updated range degraded performance values ​​that were not allowed in the original range. Updating the range of allowed performance values ​​can rise to allow any performance value of the performance flag under consideration.

[0025] Once an achievable permissible performance value is obtained for each updated range, the values ​​of the a priori parameters are updated based on the values ​​of the control parameters and the values ​​of the communication parameters that enable the permissible performance value to be achieved for each updated range. Since the constraints on at least one performance indicator (i.e., the permissible performance value) have been relaxed (by allowing degraded performance values ​​for at least one range, for example by allowing a slower convergence rate of the control algorithm of the platoon), it may be necessary to modify the values ​​of one or more a priori parameters (e.g., by reducing the speed of the vehicles in the platoon and / or by increasing the distance between the vehicles).

[0026] Then, before a change in the communication environment actually occurs (or at least before the change in the communication environment is maximal), the updated values ​​of the a priori parameters can be used to configure the geometric arrangement of the formation and / or the communication resources allocated to the formation, by, for example, transmitting the updated values ​​of the a priori parameters to the formation. Alternatively or in combination thereof, the values ​​of the control parameters and the values ​​of the communication parameters that have enabled the realization of the allowed performance values ​​for each updated range can be similarly used to configure the control algorithms and wireless links of the communication nodes of the formation, by, for example, transmitting the values ​​to the communication nodes of the formation.

[0027] In a particular embodiment, the configuration method may also include one or more of the following features considered alone or in any technically possible combination.

[0028] In a particular embodiment, the range of allowable performance values ​​for a formation parameter corresponds to a range of allowable control performance values ​​for a control algorithm used by the formation, and / or the range of allowable performance values ​​for a network parameter corresponds to a range of allowable resource usage performance values ​​for communication resources allocated for use by the formation.

[0029] In a particular embodiment, the set of parameters includes a plurality of a priori parameters including formation parameters for each formation and at least one network parameter for all formations, and searching for an updated range includes downgrading an allowable control performance value for at least one formation if it is not possible to find values ​​of control parameters and communication parameters of the formation that can achieve each range of allowable performance values.

[0030] In particular embodiments, searching for an updated scope includes first downgrading the allowed control performance value before downgrading the allowed resource usage performance value, if necessary (ie, if downgrading only the allowed control performance value is not sufficient).

[0031] In a particular embodiment, the set of parameters includes multiple a priori parameters, which include formation parameters for each formation and at least one network parameter for all formations, and if it is not possible to find values ​​of control parameters and communication parameters of the formation that can achieve each range of allowable performance values, searching for the updated range includes downgrading the allowable resource usage performance value of the formation.

[0032] In particular embodiments, searching for updated scopes includes first downgrading the allowed resource usage performance value before downgrading the allowed control performance value (if necessary).

[0033] In particular embodiments, searching for an updated range by degrading the allowed control performance value of the formation includes using the same degradation factor for all or a plurality of formations.

[0034] In a particular embodiment, the allowed control performance value of a control algorithm represents an allowed convergence speed value of said control algorithm.

[0035] In certain embodiments, the allowed resource usage performance value represents a global interference level generated by all formations when using the allocated communication resources.

[0036] In a particular embodiment, the formation parameter of each formation represents at least one of the following:

[0037] - the distance between the communication nodes of the formation,

[0038] - the shape of the geometric arrangement of the formation,

[0039] - The speed of the communication nodes of the formation.

[0040] In a particular embodiment, the allocated communication resources defined in the network parameters are any of the following:

[0041] - at least one frequency bandwidth,

[0042] - at least one spreading code or scrambling code,

[0043] - At least one time mode.

[0044] In a particular embodiment, the control parameter of the formation represents at least one weighting factor for weighting data received from other communication nodes of the formation.

[0045] In certain embodiments, the communication parameters of the formation represent at least one of the following:

[0046] - Modulation scheme,

[0047] - Channel coding scheme,

[0048] -Multiple antenna solutions,

[0049] - transmission power,

[0050] - Coverage distance,

[0051] -Packet rate.

[0052] In a particular embodiment, the configuration method is performed by a configuration server included in at least one communication node of the fleet or separate from the fleet.

[0053] According to a second aspect, the present disclosure relates to a computer program product comprising instructions, which, when executed by at least one processor, configure the at least one processor to perform a configuration method according to any one embodiment of the present disclosure.

[0054] According to a third aspect, the present disclosure relates to a configuration server, which includes at least one processor, at least one memory and at least one communication module, wherein the at least one processor is configured to execute a configuration method according to any embodiment of the present disclosure.

[0055] In particular embodiments, the configuration server is included in at least one communication node of the fleet or is separate from the fleet.

[0056] The invention will be better understood on reading the following description, given by way of example and which is in no way limiting, and made with reference to the accompanying drawings.

[0057] In the drawings, the same reference numerals denote the same or similar elements from one drawing to another. For clarity, the elements shown are not drawn to scale unless otherwise explicitly stated.

[0058] Furthermore, the order of steps shown in these figures is for illustration purposes only and is not meant to limit the present disclosure, which is applicable to the same steps performed in different orders. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] [ Figure 1 ]

[0060] Figure 1 Schematic representations of two formations corresponding to separate convoys of vehicles are shown.

[0061] [ Figure 2 ]

[0062] Figure 2 A diagram representing the main steps of a method for configuring a set of parameters for a formation is shown.

[0063] [ Figure 3 ]

[0064] Figure 3 A schematic representation of a configuration server is shown.

[0065] [ Figure 4 ]

[0066] Figure 4 Shown is a graph illustrating an example of a Pareto frontier in a multi-objective optimization problem.

[0067] [ Figure 5 ]

[0068] Figure 5 Graphs illustrating examples of different selection strategies for selecting points on the Pareto front are shown.

[0069] [ Figure 6 ]

[0070] Figure 6 A graph illustrating an example where a control performance constraint is too stringent is shown.

[0071] [ Figure 7 ]

[0072] Figure 7 A graph illustrating an example where both a control performance constraint and a resource usage performance constraint are too stringent is shown.

[0073] [ Figure 8 ]

[0074] Figure 8 Show summary Figure 2 A graph of the behavior of an example implementation of a configuration method. DETAILED DESCRIPTION

[0075] As mentioned above, the present disclosure relates to the interaction of at least two mobile formations that share at least allocated communication resources common to the at least two mobile formations for a certain period of time. In particular, modifications of the communication environment of the formations are expected to occur.

[0076] For example, the expected modification of the communication environment is due to the fact that the formations are getting closer to each other and will be in radio proximity (e.g., at least one communication node of one formation is in radio coverage of at least one communication node of another formation) and compete for the allocated common communication resources. Thus, in this case, the formations initially do not compete for the allocated common communication resources (because they are not in radio proximity to each other), but they are approaching each other, so they will soon be in radio proximity and compete for the same allocated common communication resources.

[0077] According to another example, the expected modification of the communication environment is due to the fact that at least one formation will experience a predictable degradation in the quality of its wireless link, for example, when the at least one formation moves in radio proximity with at least one other formation.

[0078] In the following, we consider in a non-limiting way that the modification of the communication environment corresponds to the fact that the formations become closer to each other and will be in the vicinity of the radios and compete for the allocated common communication resources.

[0079] Figure 1An example of such a modification of the communication environment in the case of an intersection of a first formation 10-1 and a second formation 10-2 corresponding to two fleets of vehicles is schematically represented. In the formation 10-1, each vehicle is equipped with a communication node 11, and the communication node 11 is configured to exchange data related to the control algorithm used by the formation 10-1 with other communication nodes 11 of the formation 10-1 to maintain the geometric arrangement of the formation 10-1 (for example, the relative position and speed of the communication nodes). Similarly, each vehicle of the formation 10-2 is equipped with a communication node 11, and the communication node 11 is configured to exchange data related to the control algorithm used by the formation 10-2 with other communication nodes 11 of the formation 10-2 to maintain the geometric arrangement of the formation 10-2.

[0080] In the following, when there is no need to distinguish between formations, the formations are collectively referred to as 10, and when there is a need to distinguish between formations, the formations are individually referred to as 10-i, where formation 10-i corresponds to formation 10 with index i, where 1≤i≤N, and N corresponds to the number of formations moving nearby. In the following, we will mainly focus on the case where N=2 (i.e., two formations). However, the present disclosure can also be applied to a plurality of formations 10 greater than 2.

[0081] exist Figure 1 In , the radio coverage of the formation 10-1 is denoted by RC-1, and the radio coverage of the formation 10-2 is denoted by RC-2. The radio coverage of the formation 10 corresponds to an area where messages sent by at least one communication node 11 of the formation at a given packet delivery rate (PDR) can be received and decoded. Figure 1 In part a), the formations 10-1 and 10-2 are close to each other, but initially spaced apart so that their respective radio coverages RC-1 and RC-2 do not overlap. Therefore, the communication nodes 11 from different formations 10 do not compete for the allocated common communication resources. Figure 1 In part b), formations 10-1 and 10-2 are in the vicinity, so that their respective radio coverages RC-1 and RC-2 overlap. Therefore, communication nodes 11 from different formations compete for the allocated common communication resources, so that, for example, each formation 10 may experience an increased level of packet collisions.

[0082] As described above, each formation 10 uses a control algorithm to control the geometric arrangement of the formation, preferably using a collaborative control algorithm, such as a consensus algorithm or a distributed model predictive control (DMPC) algorithm. The communication nodes 11 of the formation 10 exchange data to enable the collaborative control algorithm to control the geometric arrangement. For a convoy of vehicles, a control algorithm is used to maintain a predetermined geometric arrangement between the vehicles while moving (typically 1D with a predetermined distance and a predetermined speed between them). It is important to ensure that the control algorithm can actually control the geometric arrangement with a sufficient convergence speed to avoid, for example, collisions between vehicles. In addition, the convergence speed of the control algorithm depends on, among other things, packet losses (see, for example, [Sybis2019]). Therefore, when the radio coverage ranges RC-1 and RC-2 of the formations 10-1 and 10-2 overlap, as in Figure 1 In the lower part, the convergence speed of the control algorithm may become insufficient due to the increasing level of mutual interference (where the mutual interference level of a formation corresponds to the interference level generated by other formations).

[0083] The present disclosure relates to a configuration method 20 for configuring a set of parameters of a formation 10 in response to an expected modification of the communication environment of the formation 10. The formation 10 typically has the parameters in the set configured with predetermined values ​​before the communication environment changes.

[0084] Some of the parameters in the set are referred to as "a priori parameters" and correspond to parameters whose values ​​should be maintained when possible during changes in the communication environment. The set includes at least one a priori parameter.

[0085] Two main types of prior parameters are possible.

[0086] A first possible type of a priori parameters are formation parameters describing the geometric arrangement of the formation 10. Any formation parameters suitable for describing the geometric arrangement of the formation 10 may be used in the present disclosure. For example, the formation parameters of the formation 10 may include at least one of the following characteristics of the geometric arrangement of the formation 10:

[0087] - the distance between the communication nodes 11 of the platoon 10 (inter-vehicle distance or IVD),

[0088] - the shape of the geometric arrangement of the formation 10 (e.g. linear, square, etc.),

[0089] - The speed (eg maximum, minimum or nominal speed) of the communication nodes 11 of the formation 10.

[0090] When the a priori parameters include formation parameters, the a priori parameters include N formation parameters, one for each formation 10. If the a priori parameters include formation parameters, this means that the geometrical arrangement of the formations 10 should remain unchanged during changes in the communication environment (if possible).

[0091] A second possible type of a priori parameter is a network parameter that describes the communication resources allocated to the formation for exchanging data on a wireless link, including communication resources that are also allocated to at least one of the other formations 10 (i.e., common). The network parameter may use any format suitable for describing the communication resources allocated to the formation. For example, the allocated communication resources defined in the network parameter may be any of the following:

[0092] - at least one frequency bandwidth (e.g. a frequency channel or one or more sub-channels, one or more sub-carriers in an Orthogonal Frequency Division Multiple Access (OFDMA) system, etc.),

[0093] - at least one spreading code or scrambling code (i.e. in a code division multiple access CDMA system),

[0094] - At least one time pattern (eg, one or more time slots, etc.).

[0095] For example, a pool of communication resources (also referred to as a resource pool in the context of a 5G cellular network) available for allocation to a formation 10 may be predefined, and the network parameters of the formation define a subset of the pool initially allocated to the formation before a change in the communication environment occurs. It is assumed that at least some of the allocated communication resources are common to two or more formations 10, that is, they are allocated to two or more formations 10 at the same time. In the present disclosure, one network parameter may be defined for each formation 10, each network parameter defining the communication resources (in the pool of available communication resources) allocated to the corresponding formation 10 (at least partially shared with at least one other formation 10). However, fewer network parameters may also be considered. For example, if all formations 10 are allocated the same communication resources, a single network parameter (which may be referred to as "global" hereinafter) may be considered for all formations 10. If the a priori parameters include at least one network parameter, this means that during a change in the communication environment, the amount of communication resources allocated to the formation 10 should remain unchanged (if possible).

[0096] In the following, we consider in a non-limiting manner the use of a single (global) network parameter, such that the at least one prior parameter comprises at most one (global) network parameter.

[0097] It should be noted that in the present disclosure, the expression "parameter" may refer to a single basic (scalar) parameter, or a vector or matrix parameter comprising multiple basic parameters. In other words, the parameter comprises one or more basic parameters that may be of different types. For example, a formation parameter may include both IVD and speed. Network parameters may include, for example, one or more frequency bandwidths and one or more time modes.

[0098] When adapting to changes in the communication environment, the values ​​of the other parameters of the set (i.e., except the a priori parameters) are first reconfigured. The other parameters correspond to parameters that affect the convergence speed of the control algorithm used by the formation 10, and for each formation 10 include:

[0099] - control parameters of the control algorithm used by the formation 10 to control the geometrical arrangement,

[0100] - Communication parameters for configuring the wireless links between the communication nodes 11 of the formation 10, which are used to exchange data related to the control algorithms.

[0101] For example, in a control algorithm such as a consensus algorithm, each communication node 11 uses a control parameter corresponding to one or more weighting factors for weighting data received from neighbor communication nodes 11. Such weighting factors affect the convergence speed of the control algorithm and can be adjusted when adapting to changes in the communication environment. For example, the continuous time consensus algorithm can be summarized as:

[0102]

[0103] Among them, Θ i Represents its control information x j The set of communication nodes 11 available to communication node i at time t, and α ij In this non-limiting example, the weighting factor α can be modified by ij To modify the convergence speed. In a control algorithm such as, for example, the DMPC algorithm, the control parameters may correspond to a prediction horizon, a control horizon, and the like.

[0104] Similarly, the configuration of wireless links affects the communication performance, that is, the set Θ on these wireless links i (t) attributes (e.g., in terms of PDR, etc.), which affects the convergence speed of the control algorithm. Therefore, when adapting to changes in the communication environment, the configuration of the wireless link can be adjusted by modifying the values ​​of the communication parameters. Non-limiting examples of basic communication parameters include:

[0105] - Modulation scheme (e.g., BPSK, QPSK, 16QAM, etc.),

[0106] - channel coding scheme (e.g., type of channel coder, rate of channel coder, etc.; modulation and channel coding scheme are often collectively referred to as MCS),

[0107] -Multi-antenna scheme (e.g., number of layers, number of antennas, etc.),

[0108] - transmission power,

[0109] - Coverage distance,

[0110] - packet rate (e.g. packets per second),

[0111] - Maximum number of retransmissions, etc.

[0112] Figure 2 The main steps of a configuration method 20 for configuring a set of parameters of a formation in response to an expected change in the communication environment are schematically represented.

[0113] The configuration method 20 is executed by a configuration server 30 .

[0114] Figure 3 An exemplary embodiment of a configuration server 30 suitable for performing the configuration method 20 is schematically represented.

[0115] like Figure 3 As shown, the configuration server 30 includes one or more processors 31 and one or more memories 32. One or more processors 31 may include, for example, a central processing unit (CPU), a digital signal processor (DSP), a field-programmable gate array (FPGA), an application specific integrated circuit (ASIC), etc. One or more memories 32 may include any type of computer-readable volatile memory and computer-readable non-volatile memory (magnetic hard disk, solid state disk, optical disk, electronic memory, etc.). One or more memories 32 may store a computer program product in the form of a set of program code instructions to be executed by one or more processors 31 to implement all or part of the steps of the configuration method 20.

[0116] like Figure 3 As shown, the configuration server 30 further includes a communication module 33 for exchanging data with the fleet 10 .

[0117] In some cases, the configuration server 30 may be separate from the communication nodes 11 of the fleet 10. For example, if data is exchanged directly with the fleet 10, the communication module 33 implements at least one wireless communication protocol. If data is exchanged indirectly with the fleet 10 (i.e., via one or more intermediate communication devices), the communication module 33 may implement at least one wireless communication protocol and / or at least one wired communication protocol to exchange data with the intermediate communication device. It should be noted that the configuration server 30 may be included in multiple separate hardware devices or a single hardware device in a distributed computing architecture.

[0118] In some cases, the configuration server 30 may alternatively be included in one or more communication nodes 11 of the same fleet 10 (in a distributed computing architecture). In this case, the communication module 33 implements at least one wireless communication protocol, such as the same wireless communication protocol used by the communication nodes 11 of the fleet 10 to exchange data related to the control algorithm. If the communication module 33 implements the same wireless communication protocol used by the communication node 11, the communication module 33 may also be used by the communication module 11 to exchange data related to the control algorithm with other communication nodes 11 (i.e., the communication device 11 may include a single communication module). Moreover, the configuration server 30 is configured to exchange data with at least one communication node 11 of each other fleet 10 directly and / or indirectly via one or more other communication nodes 11 and / or one or more intermediate communication devices separate from the fleet 10.

[0119] Therefore, the configuration server 30 exchanges data with the formation 10. The data received from the formation 10 is any data required to perform the configuration method 20. For example, the received data may include the current value of the prior parameters (before the communication environment change occurs), the position and speed of the formation 10 (for example, to determine when the communication environment change will occur), etc. If the configuration server 30 is included in one or more communication nodes 11 of the formation, this means that each formation 10 can directly or indirectly, temporarily or permanently exchange data with any other formation. Depending on the nature of the communication between the formations, the formation 10 can more or less anticipate future communication environment changes in advance. If the prior parameters cannot remain unchanged, the data sent to the formation by the configuration server 30 includes the new values ​​of the prior parameters. The data sent to the formation by the configuration server 30 may also include the new values ​​of the control parameters and / or the new values ​​of the communication parameters.

[0120] like Figure 2 As shown, the configuration method 20 comprises step S20: for each a priori parameter, determining a range of allowable performance values ​​based on the corresponding a priori parameter value.

[0121] Basically, the allowed performance value is a value of a performance flag that enables the value of a priori parameter configured before the communication environment changes to be maintained during the communication environment changes.

[0122] For example, the performance indicator of the formation parameter of the formation 10 may be any control performance indicator indicating the performance of the control algorithm used by the formation 10. In a preferred embodiment, the control performance indicator indicates the convergence speed of the control algorithm, and the faster the convergence speed value, the higher the control performance value.

[0123] For example, the performance mark of the network parameters of all formations 10 can be any resource use performance mark representing the performance of the use of the allocated communication resources by the communication nodes 11 of all formations 10. In a preferred embodiment, the resource use performance mark represents the global interference level generated by all formations 10 when using the allocated communication resources. The global interference level corresponds to the interference level jointly generated by all formations 10 in the area where the formation 10 is located, and the lower the global interference level, the higher the resource use performance value. In the following, reference may be made to the resource use cost in some cases. Basically, the resource use cost makes the resource use cost value higher, the resource use performance value lower (and the global interference level is higher). For example, the resource use performance can be opposite or inversely proportional to the resource use cost, so that the resource use cost is equivalent to represent the generated global interference level, and can be represented as resource use performance by considering its opposite or its inverse ratio. For example, the resource use cost can correspond to the channel busy rate CBR defined in the 3GPP system, and the channel busy rate CBR represents the global interference level generated by the formation 10. Other examples include, for example, system load SL, average received power, etc.

[0124] In the following, the value of the formation parameter before the communication environment changes is referred to as for their respective formation parameters (where 1≤i≤N).

[0125] For formation parameters, the range of allowed control performance values ​​can be determined by their respective control performance flags. (e.g., lower bound on convergence rate) To determine the range of permissible control performance values, the function So that:

[0126] Make

[0127] Assumption Function It changes monotonically.

[0128] For example, the function A slowest convergence speed may be provided that theoretically ensures that the geometric arrangement is controlled fast enough to prevent collisions between adjacent vehicles in the formation. For example, the convergence speed may be such that the probability of the IVD becoming empty given the speed of the vehicle is below a predetermined threshold (e.g., 10 -6 or lower).

[0129] Furthermore, it is assumed that the given control performance value can be determined in advance Function This function determines the formation parameters Value

[0130] Make

[0131] Assumption Function It changes monotonically.

[0132] If the prior parameters include network parameters, then before the communication environment changes, the network parameters p net The value of is called p net，t .

[0133] For (global) network parameters, the range of allowed resource usage performance values ​​can be determined by the resource usage cost c com (e.g., CBR) com，t To determine the range of allowable control performance values, the function θ can be pre-determined net , so that:

[0134] Make c com，t =θ net (p net，t ).

[0135] In addition, it is assumed that a given value c for the resource usage cost can be determined in advance com，a Function This function determines the network parameters p net The value of p net，a :

[0136] Make

[0137] Therefore, step S20 of obtaining the allowable performance value range of each priori parameter may include:

[0138] -Based on the formation parameters The prior value of By using the function Determine the parameters of each formation Control performance The lower limit and / or

[0139] -Based on the formation parameter p net The prior value p net，t , by using the function θ net Determine the network parameter p net The resource usage cost c com The upper limit c com，t .

[0140] like Figure 2 As shown, the configuration method 20 includes step S21 of searching for values ​​of control parameters and values ​​of communication parameters of the formation 10, which enables implementation of an allowable performance value for each range in consideration of changes that have occurred in the communication environment.

[0141] As mentioned before, the performance of the control algorithm depends on the performance of the communication on the wireless link. For example, the convergence speed of the consensus algorithm is related to the spectral radius of the matrix filled by the PDR values ​​of the different links between the agents and to at least one control parameter [Pereira2011] (see also patent application EP21305452.1).

[0142] For each formation 10, we use (MCS, transmission power, packet rate, etc.) represent communication parameters that can be configured to adjust the communication performance on the wireless link. We assume in a non-limiting manner that all communication nodes 11 of the same formation 10 use communication parameters For each formation 10, we use represents a control parameter that can be configured to adjust the control performance of the control algorithm.

[0143] Therefore, for each formation 10, and / or The value of can be optimized to try and maintain the prior value of each prior parameter (i.e., the formation parameter Value and / or network parameters p net The value of p net，t ) unchanged. Since the a priori value of each a priori parameter can remain unchanged, if the performance value can be achieved within their respective allowed performance value ranges, it is assumed that the control performance flag and the resource usage performance flag are performed, which means that the optimization is performed according to the following:

[0144] - That is, the values ​​of the control parameters and the values ​​of the communication parameters enable a control performance value within the permissible range to be achieved,

[0145] -c com ≤c com，t , (i.e., -ccom ≤-c com，t , where -c com Corresponding to resource usage performance), that is, the value of the control parameter and the value of the communication parameter enable the resource usage performance value within the allowed range to be achieved.

[0146] like Figure 2 As shown, if we can find a solution that achieves each range of permissible performance values ​​without modifying the a priori value of each a priori parameter, Value and Value ( Figure 2 , each control parameter Can be configured with corresponding values And each communication parameter Can be configured with corresponding values

[0147] Furthermore, if it is not possible to find a solution that achieves each range of permissible performance values ​​without modifying the a priori value of each a priori parameter, Value and Value ( Figure 2 , the configuration method 20 includes a step S22 of searching for an updated range of permissible performance values, and determining control parameters that enable the permissible performance values ​​of each updated range in consideration of changes in the communication environment that have occurred. Values ​​and communication parameters During step S22, at least one range of permissible performance values ​​is modified to try and find control parameters that can achieve permissible performance values ​​for each range. Values ​​and / or communication parameters Modifying the range of permissible performance values ​​means including in the updated range degraded performance values ​​that were not permitted in the original range. This is done only if it is not possible to find control parameters that can achieve every range of permissible performance values. Values ​​and communication parameters When the value of , can first modify a single scope and iteratively modify other scopes. By doing so, the performance flag of the scope modified last is given priority with respect to the performance flag of the scope modified first. In other examples, multiple or all scopes etc. can also be modified simultaneously.

[0148] When the control parameters are found and communication parameters When such a value is obtained, the configuration method 20 includes step S23: updating the value of each prior parameter for the range of allowed performance values ​​that has been modified during step S22. It should be noted that in the present disclosure, all ranges obtained after step S22 are referred to as "updated ranges", but the updated ranges are not necessarily all modified with respect to the original ranges, and in some cases, it may be sufficient to modify only one range of allowed performance values ​​or one type of range of allowed performance values ​​(for example, only a range corresponding to control performance values ​​or only a range corresponding to resource usage performance values).

[0149] During step S23, based on the value of the control parameter and the values ​​of the communication parameters Update the value of at least one prior parameter, the value of the control parameter and the values ​​of the communication parameters The permissible performance values ​​for each updated range have been achieved. More specifically, the performance values ​​obtained within the updated (and modified) range (which depend on the values ​​of the control parameters and the values ​​of the communication parameters ) is used to pass through the use of function Or the corresponding function Therefore, if a performance value is obtained outside the original range of allowed performance values, the value of the associated prior parameter may need to be updated accordingly.

[0150] exist Figure 2 In the configuration method 20, the configuration method 20 comprises a step S24 in which the configuration server 30 configures the fleet 10 with the values ​​determined for the parameter set by assigning the determined values ​​to the fleet 10. This configuration step S24 is performed before the communication environment changes start, or at least before the changes are maximum.

[0151] We now provide a detailed example of the different steps of the configuration method 20. Although the configuration method 20 is preferably used to configure a set of parameters of two or more formations 10 moving in proximity, an explanation is initially provided by assuming a single formation 10, which is more convenient for introducing some concepts and notations.

[0152] {Single formation situation}

[0153] Search Control Parameters Values ​​and communication parameters The value of can be regarded as the optimization control parameter and communication parameters To make the control performance Maximize while making resource usage cost c com Minimize (i.e., minimize resource usage while keeping performance -c com maximize):

[0154]

[0155] However, these two performance indicators and are interdependent (eg, allocating more communication resources will result in better control performance most of the time), and as stated, there is no unique optimal solution to the problem.

[0156] Therefore, we are dealing with multi-objective optimization here. In the context of multi-objective optimization, we usually find The Pareto front (or Pareto front) of or -c com ) will necessarily degrade another performance indicator (see Figure 4 ).

[0157] Communication parameters This can be on bounded (e.g., maximum transmit power) and / or discrete (e.g., MCS, packet rate) support. This can also apply to control parameters And in the so-called decision (variable) space Optimization is performed on.

[0158] Therefore, dealing with multi-objective optimization requires defining criteria to select a unique "optimal" point (if any) among those points on the Pareto front. In the following, the criteria for selecting points on the Pareto front are referred to as "selection strategies".

[0159] For example, the selection strategy may include searching for a resource with a usage cost c com The solution to the minimization is:

[0160]

[0161] It should be noted that in the control performance Under the equality constraint, the resource usage cost c com Minimization will not necessarily lead to a solution to the Pareto front. Therefore, it is better to follow the inequality. In the above expression, the optimization aims to minimize the resource usage cost while ensuring that the performance value is allowed to be controlled. In other words, this selection strategy gives priority to minimizing the resource usage cost (hereinafter referred to as "communication priority selection strategy"). This method is Figure 5 is shown in part a).

[0162] Another approach could be to set priorities with respect to maximizing control performance while ensuring that resource usage cost values ​​are allowed (hereinafter referred to as "control priority selection strategy"):

[0163]

[0164] exist Figure 5 The control priority selection strategy method is shown in part b). Figure 5 Part a) (Communications Priority Selection Strategy) and Figure 5 Between the optimal points represented in part b) (control priority selection strategy), there are other optimal solutions on the Pareto front that can be obtained by considering other selection strategies.

[0165] First, we assume that the segment on the Pareto front is not empty. The selection of a single point on the Pareto front can be performed as follows:

[0166]

[0167] in, is a scalar function defined to implement the selection strategy, i.e., to select a unique solution (if any) among those on the Pareto front. In general, different scalar functions (here is and c com ) to obtain the selection of a single solution to the Pareto front, leading to a standard optimization problem. We will focus on this specific case without loss of generality. Possible solutions are the weighted sum or the Tchebycheff method [Zhang2007]. The weighted sum is well suited here with only two terms to be combined as:

[0168]

[0169] By simply setting the parameter λ = 1, we obtain a communication-first strategy, and when λ = 0, we obtain a control-first strategy. For any intermediate value, we achieve another tradeoff between the two constraints.

[0170] It may happen that the constraints cannot be fulfilled, for example due to changes in the communication environment which introduce transmission impairments. ( Figure 2 Reference S211), such as Figure 6 In this case, one possibility is to allow the a priori values ​​of the formation parameters to be modified. This can be done by constraining the resource usage cost c to be unchanged. com，t The control performance is maximized by:

[0171]

[0172] This corresponds to updating the range of allowed control performance values ​​by allowing essentially any control performance value given the communication cost constraint. Then, from the value and Derive the values ​​of formation parameters, where the values and The updated range c has been enabled com ≤c com，t (not modified during step S22) and The allowed performance value in (modified during step S22):

[0173]

[0174] Figure 7 It shows that due to the resource usage cost constraint c com，t Another example of a solution that is too strict. In this example, the constraints on control performance are relaxed is not enough. The only solution is to also relax the constraint on resource usage cost c com，t . Dealing with the communication priority selection strategy, the solution is to also reduce the resource usage cost constraint so as to achieve the first non-zero control performance value:

[0175]

[0176] In this example, this corresponds to updating the range of allowed control performance values ​​and the range of allowed resource usage cost values ​​by allowing substantially any control performance value and any resource usage cost value. Figure 7 In the example, for all performance values ​​outside the original range of allowed performance values, the solution is implemented The values ​​of the formation parameters and the network parameters are derived from the values and These values ​​are exported and It has been possible to update the scope c com ≤∞ (modified during step S22) and The allowed performance value is achieved in (modified during step S22):

[0177]

[0178] More generally, when it is not possible to find values ​​of the control parameters and values ​​of the communication parameters that enable performance values ​​to be achieved within the original range of allowed performance values, four main cases are possible:

[0179] -Control performance constraints It cannot be realized, but the resource usage cost constraint c com，t It can be achieved.

[0180] - Resource usage cost constraint c com，t Not achievable, but control performance constraints It can be achieved.

[0181] - Resource usage cost constraint c com，tand control performance constraints None of them can be achieved.

[0182] - Resource usage cost constraint c com，t and control performance constraints Both are possible, but not simultaneously.

[0183] Therefore, during step S22, it is necessary to reduce the control performance constraints according to the predetermined criteria Or add resource usage cost constraint c com，t Or both. Hereinafter, the criteria applied to selectively downgrade the range of allowed performance values ​​are referred to as "prioritization policies." For example, for a communication priority prioritization policy, the range of allowed control performance values ​​is first modified (downgraded). For a control priority prioritization policy, the range of allowed resource usage performance values ​​is first modified (downgraded).

[0184] This can be summarized as:

[0185]

[0186] in, is used to implement a selection strategy (which may be different from ), and G ↓ (.) Defines how the range of allowed performance values ​​is updated during step S22.

[0187] {General multi-objective problem}

[0188] The general multi-objective problem can be expressed as:

[0189]

[0190] Where F(x)=(f 1 (x), …, f m (x)) is the objective function, (If m>1, it is a multi-objective optimization), and the decision variable x is a , ie, the decision space. is a set of decision variables, and Ω is a set of constraints and is included in the set middle.

[0191] Since F(x) is a vector of at least two dimensions (m>1), we do not have a total order, i.e., for all pairs of vectors F(x), we do not have a higher (or equal) value. When F(x) is a vector, we have to define our meaning as max(F(x)), and we consider this maximum in the Pareto sense. The maximum of such a vector function is such that if a particular vector is considered as a maximum, it is impossible to increase one of the components without decreasing another. In this sense, we generally have many solutions: the set of such solutions is what we call the Pareto front. The Pareto frontier defined by Therefore included in middle.

[0192] We can consider more explicit constraints by:

[0193]

[0194] Where G(x) = [g j (x), j = 1...C] t and C is the number of scalar constraints. g j (x) is defined as the following combinatorial function

[0195]

[0196] Therefore, the general multi-objective optimization problem leads to If the algorithm leads to several solutions, or if we can convert the multi-objective optimization problem into a single-objective optimization problem (e.g., communication-first versus control-first), then we can arbitrarily choose a particular solution.

[0197] Here, we define the function Right now, Right now This function can be parameterized via explicit arguments.

[0198] We consider the following general approach, which covers all optimization methods based on the max-min optimization approach:

[0199]

[0200] The function min is a vector function of dimension m′≤m (m≥N, where N is the number of formations) The minimum value of each component of . More specifically, is a single function f i(x) is a linear combination of functions (matrix A of size m′×m), f i (x) can be obtained by constraining f t，i Parameterized:

[0201] [Mathematical formula 1]

[0202]

[0203] {Multiple formations}

[0204] For multiple formations 10, we also face a multi-objective optimization problem. If we consider the communication priority selection strategy, an exemplary method is to make the resource usage cost C of all formations 10 under the control performance inequality constraint of each formation com Minimize:

[0205]

[0206] That is, the allowed range corresponds to the i=1...N and c com ≤c com，t .

[0207] The above optimization focuses on minimizing resource usage costs. Similar to the case of a single formation, it may be relevant to focus on maximizing control performance, such as the following:

[0208]

[0209] in, is a scalar function defined, for example, according to a weighted sum or Tchebycheff method.

[0210] Another possible and advantageous approach is to establish a compromise between the two constraints using the same method as for a single formation:

[0211]

[0212] When some formations 10 or all formations 10 cannot achieve control performance within their respective allowable control performance value ranges and / or formations 10 cannot achieve resource use cost values ​​within their respective allowable resource use cost value ranges ( Figure 2 In the case of reference numeral S211 in FIG. 1 , we can again rely on the update of the allowed range. Again, depending on whether all constraint regions (allowed ranges) overlap with the Pareto front (but do not overlap with each other), we can satisfy two cases. In the case where the range of allowed resource usage costs overlaps with the Pareto front, the solution is to maximize the achievable control performance, for example:

[0213]

[0214] As an example, it is relevant to ensure that all formations 10 suffer a proportional degradation of their control performance:

[0215]

[0216] Therefore, the allowable range of resource usage cost C com ≤C com，t unchanged, while the permissible range of control performance values ​​is determined by allowing any control performance value To modify. Since the allowed range of control performance values ​​is modified (degraded) first, we have a communication priority prioritization strategy here. Once the allowed range of control performance values ​​is modified, we use a control priority selection strategy under the fairness (proportional) control performance degradation constraint.

[0217] Can be used Or by using different functions to optimize, such as:

[0218] For any j∈[1,N].

[0219] Instead of using fairness (proportional) control performance degradation constraints, the fairness degradation of formation parameters can be replaced by using, for example, the following fairness constraints:

[0220]

[0221] In general, the problem can be rewritten as:

[0222]

[0223] in,

[0224] It may happen that some equalities cannot be achieved on the Pareto front. From this point of view, it is better to rely on inequality constraints:

[0225]

[0226] The two previous cases can be considered a max-min problem, since it reduces N degrees of freedom to one:

[0227]

[0228] in:

[0229]

[0230] This method can be further extended to the case where the range of resource usage cost values ​​does not overlap with the Pareto frontier. In this case, it is not only necessary to control the resource usage cost value at a lower level than the original control performance constraint. The control performance level needs to be higher than the original communication cost constraint C com，t The general problem can be solved as follows:

[0231]

[0232] A possible solution could be to request that all performance values ​​(control and resource usage) be degraded proportionally, i.e.

[0233]

[0234] Among them, for example:

[0235]

[0236] In general, the problem can be summarized as finding the parameters Make the performance value satisfy the constraints (the range of allowed performance values):

[0237]

[0238] According to any multi-objective optimization algorithm, a unique solution (if any) is generated on the Pareto front. In the case that the target task cannot be achieved, the algorithm searches for an achievable performance value on the Pareto front according to a predetermined selection strategy (such as a communication priority selection strategy or a control priority selection strategy) and a predetermined priority sorting strategy. =1...N,c com，a (ie, by updating the allowed range).

[0239] Therefore, during step S20, the value i=1...N and c com，t To obtain the range of allowable performance values.

[0240] During step S21, the control parameters are optimized according to a given selection strategy (such as communication priority, control priority, fairness (ie, proportional degradation), etc.). and communication parameters To meet control performance constraints and resource usage cost constraints:

[0241]

[0242] in, is a scalar function defined to implement the selection strategy (and Ω ↑ ), that is, to select a unique solution (if any) among those on the Pareto front.

[0243] If it is not possible to achieve performance values ​​within the permitted range ( Figure 2, then during step S22, the allowed range may be updated (by modifying at least one allowed range) to try and find a value of the control parameter and a value of the communication parameter that enables a performance value within the updated allowed range to be achieved:

[0244]

[0245] Among them, G ↓ (.) is defined as reducing the constraints on one or more allowed ranges (in other words, G ↓ a function vector defining how to update the range of allowed performance values ​​during step S22) to select a non-empty segment of the Pareto front, and is a scalar function defined to implement the selection strategy. Step S22 may consist of several steps during which G ↓ can be modified (e.g., by increasing the number of modified ranges of allowed values, i.e., by increasing the number of relaxed constraints) and possibly, can be modified until a solution is found. A detailed example is provided below.

[0246] When a solution is found, the values ​​of the control parameters and the values ​​of the communication parameters that have been such that a performance value in the updated allowed range is achieved provide the performance value achieved for the selected solution. and / or c com，a :

[0247]

[0248] Then, during step S23, the values ​​of the formation parameters and / or the values ​​of the network parameters may be updated:

[0249]

[0250] The above behavior Figure 8 In summary.

[0251] {Realization of Fairness Solution}

[0252] From an application point of view, a solution that guarantees a proportional degradation of the formation parameters with respect to their a priori values ​​appears very relevant, but is non-limiting, since it provides a kind of fairness among all formations 10. This optimization problem can be solved using the well-known max-min optimization method:

[0253]

[0254] For example, or And if we want, α iis chosen to vary the weight of each formation, and may be chosen to be all equal to 1. Note that the proposed method also operates if the communication parameters remain constant. The algorithm can still be used to reduce the formation parameters to values ​​compatible with the radio environment.

[0255] {Separation of formation parameters in two groups}

[0256] In some cases, the formation parameters can be divided into two groups: are the formation parameters that are not updated for some reason, and another set of functions that effectively control performance can be updated, i.e., Notice, Can depend on the obtained This means is decomposed into two functions

[0257]

[0258] function Can be is the (maximum) speed and For example, in the case where two convoys can communicate well in advance using a cellular network, the two convoys can adjust their IVD and speed before crossing (i.e., with a known convergence rate). However, if the convoys can only communicate in a distributed mode, they do not have much time to adapt their configurations and may be more suitable to simply reduce speed without increasing IVD (an operation that may not be done before the convoys cross each other). It may also be undesirable to change the IVD, because it may be needed to increase it again later, but there may not be enough space due to traffic congestion. We can perform targeted The same type of decomposition.

[0259] {Generalization of the proposed optimization algorithm}

[0260] We now provide some detailed examples of steps S21 and S22 in the framework of the above section {General multi-objective problem}. We present a first example of possible functions that can be used to transform a multi-objective optimization problem into a single-objective optimization problem. We consider N formations such that we have at the input the maximum range of N+1 ranges of allowed performance ranges defined by the corresponding performance objective values:

[0261] -Control performance constraints:

[0262] - Resource usage cost constraint (if any): c com，t

[0263] We now give the function h i, i=1...N and possible examples of matrix A (see equation [Math. 1] above), which can be used during steps S21 and S22.

[0264] For example, a constraint (a range of allowed property values) can be expressed as follows:

[0265]

[0266] For example, we can consider h i The following function:

[0267] - Case 1: Relative objective function with respect to basic objectives (formation and network):

[0268]

[0269] - Case 2: Relative objective function with respect to the target (controlling the opposite of performance and resource usage cost):

[0270]

[0271] - Case 3: Objective function (formation and network):

[0272]

[0273] - Case 4: Objective function (controls the opposite of performance and resource usage cost):

[0274]

[0275] For example, we can consider the following definition of the matrix A:

[0276] - Case 1 (where, for any i=1...N+1, α i can be equal to 1):

[0277]

[0278] - Case 2 (where, for any i=1...N+1, α i can be equal to 1):

[0279]

[0280] - Case 3 (where, for any i=1...N+1, α i can be equal to 1):

[0281]

[0282] By considering the original range of allowed performance values ​​and a given selection strategy (communication priority, i.e., minimizing resource usage cost, control priority, i.e., maximizing control performance, fairness strategy, etc.) for selecting a solution on the Pareto front (if any), the function h can be used during step S21 i and these examples for the matrix A.

[0283] If no solution on the Pareto front exists when considering the initial range of allowed performance values, some constraints on the performance values ​​need to be relaxed. During step S22, a different function h than during step S21 may be used. i and matrix A, and a different selection strategy than during step S21 may also be used.

[0284] We now elaborate on three different examples of step S22 for different prioritization strategies (regarding the priority of the range of allowed performance values) and different selection strategies (regarding the selection of solutions on the Pareto front).

[0285] {Step S22 of utilizing communication priority prioritization strategy}

[0286] Even though the matrix A corresponds to case 1 above during step S21, for step S22, we consider case 2 or 3 of the matrix A.

[0287] It should be noted that when considering the communication priority selection strategy during step S21, we usually set λ=1 (in order to maximize the resource usage performance). However, in this example, we use a prioritization strategy, which means that we initially do not change the range of allowed resource usage performance values, and we only modify (downgrade) all or part of the range of allowed control performance values. Since the range of allowed resource usage performance values ​​is not modified, in this example, we apply the control priority selection strategy to maximize the control performance in the range of allowed modification (downgrading) of control performance values. Therefore, we initially set λ=0, and the range of allowed modification of control performance values ​​can allow any positive value of control performance.

[0288] If no solution is found even by allowing any positive value of the control performance of the control algorithm of the formation 10, we also modify (downgrade) the range of allowed resource usage performance values. For example, we can further allow any resource usage performance value. Then, we can set λ=1 and apply the communication priority selection strategy to search for a solution on the Pareto front that maximizes the resource usage performance (i.e., it minimizes the resource usage cost).

[0289] Therefore, in this example, the range of allowed control performance values ​​is first downgraded, and only if no solution is found by downgrading the range of allowed control performance values ​​alone, the range of allowed resource usage performance values ​​is downgraded.Thus, a solution is found during step S22 by performing at most two optimization phases.

[0290] {Step S22 of using control priority prioritization strategy}

[0291] Even if the matrix A corresponds to Case 1 above during step S21, for step S22, we consider Case 2 or Case 3 of the matrix A.

[0292] It should be noted that when considering the control priority selection strategy during step S21 , we usually set λ=0 (in order to maximize the control performance).

[0293] However, in this example, we use a prioritization strategy, which means that we do not initially change the range of allowed control performance values, and we only modify (downgrade) the range of allowed resource usage performance values. Since the range of allowed control performance values ​​is not modified, in this example, we apply a communication priority selection strategy to maximize the resource usage performance in the range of allowed modification (downgrading) of resource usage performance values. Therefore, we initially set λ=1 and the range of allowed modification of resource usage performance values ​​can allow any positive value of resource usage performance.

[0294] If no solution is found even by allowing any value of the resource usage performance of the fleet 10, we also modify (downgrade) the range of allowed control performance values. For example, we can further allow any control performance value. Then, we can set λ=0 and apply the control priority selection strategy to search for a solution that maximizes the control performance on the Pareto front.

[0295] Therefore, in this example, the range of permissible resource usage performance values ​​is first degraded, and then the range of permissible control performance values ​​is degraded only if no solution is found by only degrading the range of permissible resource usage performance values. Thus, a solution is found during step S22 by performing at most two optimization phases.

[0296] {Step S22 with no constraints priority}

[0297] Even if the function h i During step S21 corresponding to case 3 or 4, for step S22 we also preferably consider the function h iCase 1 or Case 2. During step S22, for matrix A, Case 1, Case 2 or Case 3 is possible. For complete fairness (i.e., constraints have no priority), a matrix A corresponding to an identity matrix of size N+1 may be considered during step S22. Thus, the goal is to search for a matrix that tends to apply the same degradation factor to all ranges of allowed performance values ​​(i.e., about the target value i=1…N+1 and C com，t ). Consider that the matrix A is the identity matrix N+1, and for the function h i In case 1, the maximum minimum will tend to find a solution such that:

[0298]

[0299] This maximizes the degradation factor p while allowing any performance value to control performance and resource usage performance.

[0300] Thus, in this example, all ranges of allowed performance values ​​are degraded simultaneously.Thus, a solution is found by performing a single optimization phase during step S22.

[0301] We now describe alternative and non-limiting embodiments for some aspects of the present disclosure.

[0302] {Optimization using reference performance model}

[0303] It is better to rely on analytical models to evaluate the control performance in the optimization phase (steps S21 and S22). The problem is that the control algorithms used in practice are usually complex, making it difficult to obtain analytical expressions of their control performance in some cases. The solution is then to rely on a control performance model associated with a simple control algorithm, and another control algorithm is used to actually control the formation 10. The goal is then, for example, to Applying margins (eg, verified through simulation) can account for differences in control algorithms.

[0304] {Separate optimization of communication parameters and control parameters}

[0305] As mentioned above, the control performance of the control algorithm of each formation 10 depends on the communication parameters and the control parameters according to the following type of relationship:

[0306]

[0307] In addition to what has been introduced, control performance depends on communication performance. Therefore, for optimization, we may need a method to calculate the communication performance function based on the communication parameters of all communication nodes 11 (such as a PDR curve). This can be done by using a system simulator at the expense of high processing complexity, or by using a look-up table (LUT) at the expense of high storage complexity. The most appropriate solution is to rely on an analytical transmission model that will be derived from the parameters and other parameters describing the characteristics of the radio channel provide the communication performance required to calculate the control performance. The problem with this last approach is that it only allows very simple scenarios (e.g. ignoring packet conflicts or having a very simplified description of packet conflicts). For example, one can use the model in [Gonzalez2019], which is semi-analytical and uses analytical expressions, but also uses relationships derived empirically based on system-level simulations. It considers packet conflicts, but in a simplified way and considers the simplest homogeneous model of this packet conflict (infinite regular one-dimensional vehicle positions and the same communication parameters for all vehicles).

[0308] Similarly, we may also need a method to calculate the control performance function based on the communication performance function and the control parameters method. Just like the communication performance function, a variety of methods can be used for this purpose, such as Monte-Carlo simulations. It is easier to rely on analytical expressions, such as the lower bound on the convergence rate of the first-order static consensus algorithm in [Pereira2011]. In this case, we can subtract the PDR matrix from the PDR curve, the PDR matrix for a given topology and number of vehicles in the formation (and possibly the correlation matrix from the correlation characteristics) from the homogeneity within the platoon (e.g., the same PDR for all vehicles), and then we can build the (symmetric) matrix and finally compute its spectral radius.

[0309] Another solution could be a Monte Carlo simulation of the control algorithm to obtain the relationship between the communication performance curve and the control performance. It allows considering all control algorithms, scenario types and control performance indicator types (e.g., probability of emergency outage), not just those for which we have analytical versions (e.g., lower bounds on previous true performance).

[0310] {Use of a single transmission model for multiple formations}

[0311] The proposed technique is applicable to any kind of formation 10. However, it is easier to implement in the case of homogeneous formations, i.e., formations organized according to a regular geometrical arrangement allowing all communication nodes 11 to use the same communication parameters and control parameters (except for a possible leader having parameters different from those of the followers). The homogeneous case allows to reduce the number of parameters to optimize (almost divided by N) and to rely on simplified but still accurate models to evaluate the communication performance.

[0312] Dealing with multiple fleets,10 of vehicles interacting with each other is usually homogeneous individually but not globally. Designing a communication model compatible with this case is obviously more complex. We propose here a method that allows the use of a simple homogeneous model even in the non-homogeneous case. It is assumed that a homogeneous communication model is available as a LUT (which can be obtained from system-level simulations or from a simplified (quasi-)analytical model).

[0313] The communication performance can be measured in terms of PDR, which depends on the distance between the transmitter and the receiver (since we consider homogeneity, it does not depend on the specific transmitter and receiver locations). It is also possible to consider the correlation between these PDRs, but at the cost of significantly increasing the complexity.

[0314] Regarding communication parameters, here we consider PDR to be related to at least coverage (which is related to transmission power, average path gain power, MCS frame error rate (FER) to signal to noise ratio (SNR) and a reference FER value (which may be a sensing threshold and / or half-duplex errors), and corresponds to coverage and packet rate in the absence of shadowing and packet collisions. If the transmission power is constant (e.g., the maximum possible), coverage is not required. The distance to vehicle spacing or to coverage can be normalized.

[0315] For the non-limiting example considered here, we consider the coverage Cov and the packet rate PR as the basic communication parameters (i.e., x com =(Cov, PR) and considering PDR (the curve of PDR) as the communication performance, that is, the homogeneous model (therefore, for a formation with homogeneous formation internal parameters) is the function:

[0316]

[0317] in, is from arrive is a set of continuous functions of (more precisely, decreasing to 0 as the distance approaches infinity).

[0318] This PDR should represent the PDR without half-duplex errors. To obtain the final PDR, we consider PDR(d; Cov, PR) × (1-HDER(PR)), where HDER(PR) is a function of the half-duplex error rate that does not depend on the distance or coverage communication basic parameters.

[0319] To obtain a heterogeneous model from a homogeneous model, we can make the following assumptions:

[0320] - When d is the distance normalized to the coverage distance (i.e., divided by the coverage distance), we consider each formation to have the same PDR model PDR(d; Cov Glob , P.R. Glob ),

[0321] -Cov Glob and PR Glod is calculated for each formation 10 for all i and possible spacing Δ i (IVD) A given function of the values ​​of (and those values ​​that may be associated with additional non-convoy vehicles in an average manner).

[0322] Here we describe the (The spacing Δ is not considered here i )Get Cov Glob and PR Glob Two possible approaches:

[0323] -Method 1:

[0324]

[0325] -Method 2:

[0326]

[0327] Both approaches result in the same system load, which is linear (the sum of the system loads is the sum of the system loads of each formation):

[0328]

[0329] Method 1 is accurate when all formations 10 have the same communication parameters, but overestimates the PDR for very different communication parameters.

[0330] On the other hand, method 2 is coherent for very different communication parameters (from the case of significantly different control performance for two convoys of the same size), but is not accurate in the case of homogeneous identical formations (ie, identical communication parameters).

[0331] {Use two control performance functions}

[0332] When we do not have a simple formula for the control performance of a given control algorithm, we can use another control performance indicator that we believe is close to the performance of the control algorithm we use. In this case, we can use the given control performance indicator for the optimization in steps S21 and S22 (for reasons of computational complexity) and use another control performance indicator that uses the obtained communication (and possibly control) parameters and that is more realistic (for example, this control performance indicator can be calculated by Monte Carlo simulation). Then, we can use the control performance indicator with fixed The control performance flag of the second type (i.e. fixed communication parameters) is used to optimize the control parameters associated with the control performance flag, i.e., for each formation, the control performance flag is maximized on its own control parameters. The function for updating the formation parameters (in step S23) should use the control performance flag of the second type instead of the control performance flag used during optimization (steps S21 and S22).

[0333] It should be emphasized that the present disclosure is not limited to the above exemplary embodiments. Modifications of the above exemplary embodiments are also within the scope of the present invention.

[0334] For example, the above exemplary embodiments have been provided by focusing on a single (global) network parameter. However, as mentioned above, in some cases, more than one network parameter may also be considered, such as one network parameter per formation 10. The same methods discussed above for handling multiple formation parameters may be similarly applied to handling multiple network parameters.

[0335] {References}

[0336] [Chen2005] Yang Quan Chen and Zhongmin Wang, "Formation control: a review and a new consideration", 2005 IEEE / RSJ International Conference on Intelligent Robots and Systems, 2005, pp.3181-3186, doi:10.1109 / IROS.2005.1545539

[0337] [Sybis2019] M. Sybis et al., "Communication Aspects of a ModifiedCooperative Adaptive Cruise Control Algorithm", IEEE Transactions on Intelligent Transportation Systems, Vol. 20, No. 12, pp. 4513-4523, December 2019, doi:10.1109 / TITS.2018.2886883

[0338] [Pereira2011] S. Silva Pereira and A. Pages-Zamora, "Consensus in Correlated Random Wireless Sensor Networks", IEEE Transactions on Signal Processing, vol. 59, no. 12, pp. 6279–6284, December 2011, doi:10.1109 / TSP.2011.2166552

[0339] [Zhang2007] Q. Zhang and H. Li, "MOEA / D: A Multiobjective Evolutionary Algorithm Based on Decomposition", IEEE Transactions on Evolutionary Computation, Vol. 11, No. 6, pp. 712-731, December 2007, doi:10.1109 / TEVC.2007.892759

[0340] [Gonzales2019] Manuel Gonzalez-Martín, Miguel Sepulcre, Rafael Molina-Masegosa, and Javier Gozalvez, "Analytical Models of the Performance of C-V2X Mode4 Vehicular Communications," IEEE Transactions on Vehicular Technology, Vol. 68, No. 2, February 2019

Claims

1. A computer-implemented method for configuring values ​​of a set of parameters of at least two mobile formations, each formation comprising a plurality of corresponding communication nodes, wherein: The set of parameters comprises at least one a priori parameter, wherein each a priori parameter corresponds to a formation parameter of each formation or a network parameter of the formation, wherein the formation parameter of the formation defines a geometric arrangement of the communication nodes of the formation, and the network parameter of the formation defines communication resources allocated to the formation for exchanging data on a wireless link, wherein each a priori parameter has a predetermined value, and a predetermined change of the communication environment of the formation is about to occur, wherein, for each formation, the set of parameters further comprises: - control parameters of a control algorithm used by said formation to maintain said geometrical arrangement, - communication parameters for configuring the wireless links between the communication nodes of the formation, the wireless links being used to exchange data related to the control algorithm, The configuration method comprises the following steps: - for each a priori parameter: obtaining a range of permissible performance values ​​based on the corresponding a priori parameter value, - when a change in the communication environment of the formation is considered to have occurred, searching for the values ​​of the control parameters and the communication parameters of the formation that can achieve an allowable performance value within each range, - if it is not possible to achieve the permissible performance value for each range, searching for an updated range of permissible performance values ​​for which, when taking into account that the change in the communication environment has occurred, it is possible to determine the value of the control parameter and the value of the communication parameter that enable the permissible performance value for each updated range to be achieved, - determining an updated value of said at least one a priori parameter based on the values ​​of said control parameter and of said communication parameter enabling the achievement of an allowed performance value for each updated range of allowed performance values.

2. The method according to claim 1, wherein: The range of allowable performance values ​​of formation parameters corresponds to the range of allowable control performance values ​​of the control algorithm used by the formation, and / or the range of allowable performance values ​​of network parameters corresponds to the range of allowable resource usage performance values ​​for communication resources allocated for use by the formation.

3. The method according to claim 2, wherein: The set of parameters includes multiple a priori parameters, which include formation parameters of each formation and at least one network parameter of all formations, and if it is not possible to find values ​​of the control parameters and values ​​of the communication parameters of the formation that can achieve the allowable performance values ​​of each range, searching for the updated range includes downgrading the allowable control performance value of at least one formation.

4. The method according to claim 3, wherein: Searching for an update scope includes first downgrading the allowed control performance value, if necessary, before downgrading the allowed resource usage performance value.

5. The method according to claim 2, wherein: The set of parameters includes multiple a priori parameters, which include formation parameters of each formation and at least one network parameter of all formations, and if it is not possible to find values ​​of the control parameters and values ​​of the communication parameters of the formation that can achieve each range of allowable performance values, searching for the updated range includes downgrading the allowable resource usage performance value of the formation.

6. The method according to claim 5, wherein: Searching for an update scope includes first downgrading the allowed resource usage performance value before downgrading the allowed control performance value, if necessary.

7. The method according to any one of claims 3, 4 and 6, wherein: Searching for an updated range by degrading the permissible control performance value of the formation includes using the same degradation factor for all or a plurality of the formations.

8. The method according to any one of claims 2 to 7, wherein: The allowable control performance value of the control algorithm represents an allowable convergence speed value of the control algorithm.

9. The method according to any one of claims 2 to 8, wherein: The allowed resource usage performance value represents a global interference level generated by all formations when using the allocated communication resources.

10. The method according to any one of claims 1 to 9, wherein: The formation parameter of each formation represents at least one of the following: - the distance between the communication nodes of the formation, - the shape of said geometrical arrangement of said formations, - The speed of the communication nodes of the formation.

11. The method according to any one of claims 1 to 10, wherein: The allocated communication resources included in the network parameters are any of the following: - at least one frequency bandwidth, - at least one spreading code or scrambling code, - At least one time mode.

12. The method according to any one of claims 1 to 11, the method being performed by a configuration server, the configuration server being included in at least one communication node of a formation or being separate from the formation.

13. A computer program product comprising instructions which, when executed by at least one processor, configure the at least one processor to perform the method according to any one of claims 1 to 12.

14. A configuration server, comprising at least one processor, at least one memory and at least one communication module, wherein: The at least one processor is configured to perform the method according to any one of claims 1 to 12.

15. The configuration server according to claim 14, the configuration server being included in at least one communication node of a fleet or being separate from the fleet.

Images