Method for transmitting information in a distributed system

Abstract

The invention relates to a method (100) for transmitting information in a distributed system, comprising - estimating (101) information available at at least one sink node of the distributed system, - evaluating (102) a relative information gain (30) of information (40) available at one or more source nodes of the distributed system in relation to the estimated (101) information available at the at least one sink node on the basis of a function, wherein the function depends on at least one probability distribution of the information (40) available at the at least one sink node and / or at the at least one source node, - requesting (103) an allocation of resources in a network on the basis of the evaluated (102) relative information gain (30) in order to prioritise the transmission of the information (40) and thus to maximise an expected information gain for the at least one sink node.

Description

[0001] R. 415547 - 1 -Description Title Method for Transmitting Information in a Distributed System The invention relates to a method for transmitting information in a distributed system. Furthermore, the invention relates to a computer program, a device, a platform, a storage medium, an evaluation function, and a message for this purpose. Prior Art Distributed systems, encompassing various applications such as vehicle networks, industrial automation, and sensor networks, rely on data exchange between their components. This can occur within individual vehicles, between connected autonomous transport vehicles (AGVs) in an industrial complex, or across large areas including forests or oceans. However, these systems are limited by the amount of data that can be shared due to factors such as limited communication resources and energy constraints.In the context of road traffic automation, reliable environmental models are essential so that autonomous vehicles can perceive their surroundings and recognize potential hazards. To achieve this, data from on-board sensors are aggregated and integrated into the vehicle's environmental model. Vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communication, collectively referred to as vehicle-to-X (V2X) communication, have seen significant developments in recent years. 2 -It has gained attention. Through V2X communication, vehicles can improve their environmental models by supplementing data from their on-board sensors with incoming messages from other vehicles or fixed infrastructure sensors. Collective Perception, a state-of-the-art V2X service, aims to share information about the current driving environment via V2X communication between connected nodes. To this end, the equipped nodes exchange so-called Collective Perception Messages (CPMs), which contain abstract descriptions of the detected objects and properties of the perceptual area. A typical scenario where collective perception is useful involves, for example, highway on-ramps, where merging vehicles share detected objects with vehicles already on the highway. CPMs primarily contain information about an ego vehicle, its on-board sensors, a list of detected objects, and perceptual areas.Collective Perception reduces environmental uncertainty by enabling connected nodes to perceive a greater number of objects or improve the quality of object data, as well as by exchanging occupancy information across perception areas and improving their perception quality, also known as perception area confidence. The Intelligent Transport Systems (ITS) Committee of the European Telecommunications Standards Institute (ETSI) has standardized an initial version of the Collective Perception Service (CPS) for testing purposes and is currently preparing it for commercial deployment. Key associations and alliances have raised concerns about the usability of the first version of CPS and are advocating for a simplification of the currently very complex rules for message generation.C-ROADS, an association of European road operators currently specifying a communication profile for the CPS used by the infrastructure, is even considering amending the rules to R. 415547. 3 -To circumvent message generation, it is advisable to avoid this if it is not significantly simplified in the final version of the standard. Reasons for this include the difficult implementation of the protocols and the potential performance inefficiencies of the service resulting from its high complexity and, in some cases, counterproductive functionalities. Furthermore, this complexity hinders the efficient use of the scarce communication resources allocated to ITS applications. To optimize the value of shared information, considering physical and energy constraints, mechanisms are needed that maximize the sharing of valuable information. Ideally, these mechanisms should operate decentrally, without requiring the exchange of additional administrative information to coordinate the system, thereby minimizing channel resource consumption.Disclosure of the Invention: According to aspects of the invention, a method with the features of claim 1, a data processing device with the features of claim 12, an evaluation function with the features of claim 13, a platform with the features of claim 14, a computer program with the features of claim 15, a computer-readable storage medium with the features of claim 16, and a message with the features of claim 17 are provided. Further features and details of the invention will become apparent from the respective dependent claims, the description, and the drawings. Features and details described in the context of the method according to the invention also apply to the computer program, the data processing device, the evaluation function, and the message according to the invention, and vice versa.According to one aspect of the invention, a method for transferring information in a distributed system, wherein the method comprises: -estimating available information at at least one sink node of the distributed system, R. 415547 -. 4 -- Evaluating the relative information gain of information available at one or more source nodes of the distributed system in relation to the estimated information available at the at least one sink node, based on a function wherein the function depends on at least one probability distribution of the information available at the at least one sink node and / or at the one or more source nodes, wherein the function depends in particular on covariances of the available information; - Requesting an allocation, preferably the allocation, of resources in a network based on the evaluated relative information gain in order to prioritize the transmission of the information and thus maximize an expected information gain for the at least one sink node. In other words, in a first step, information available at at least one sink node of the distributed system is estimated.The probability distribution can represent a covariance of the information. It is possible to optimize information transfer within a distributed system by prioritizing data transmission based on its potential value. This enables more efficient use of network resources by considering both the information available at the receiving nodes and the potential gain that the source nodes can offer. By using probability distributions and evaluating relative information gains, the system can determine which information is most likely to be useful to the receiving nodes. This inventive approach has the advantage of aiming to maximize the overall expected information gain for the distributed system.A particular advantage is the additional consideration of the estimated resources required for the transmission of each evaluated piece of information. This allows for the optimization of information gain across the entire network, even with heterogeneous resource requirements and information gains from the available information. Furthermore, the method may also include the estimation of available information at the at least one sink node only R. 415547 -. 5 -The estimation is performed once, based on a specific time period and / or based on the condition that the at least one sink node of the distributed system receives the prioritized information only from one source node of the distributed system. If the defined time period ends or the condition is no longer met, a new estimation is performed by the single source node. It is also possible for the procedure to further include the periodic and / or event-based estimation of available information at the at least one sink node, where the event depends on a change in the environment of the distributed system and / or a change in network resources.Estimates according to the present invention, particularly estimates by a source node or entity, are understood to mean making a reasoned judgment or calculation regarding a value, quantity, or result based on available information, data, observations, or experience. Estimates involve a degree of uncertainty but are based on the analysis of known information. Estimates are used when precise data are unavailable but an approximate value is required. The information available at the at least one sink node can be estimated based on: - the information received from one or more source nodes and / or - the information transmitted by the one or more source nodes.More precisely, the information available at at least one sink node can be the sum, union, or combination of the information received from one or more source nodes and / or the information transmitted by the one or more source nodes. R. 415547-. 6 -Alternatively, the information available at at least one sink node can be estimated as information transmitted by the one or more source nodes. It is also possible that the distributed system, or more precisely, a system structure of the distributed system, is configured as a centralized system with a central entity, the procedure further comprising receiving and aggregating the available information transmitted by one or more source nodes by the central entity. Alternatively, the distributed system, or more precisely, a system structure of the distributed system, is configured as a decentralized cluster-based network topology with a cluster head, the procedure further comprising receiving and aggregating the available information transmitted by one or more source nodes by the cluster head.The system structure can be defined either as a centralized system with a central entity or as a decentralized, cluster-based network topology, the latter comprising multiple clusters, each with a head node that aggregates information from the source nodes. This flexibility in defining the system structure allows for adaptation to different scenarios and environments, potentially leading to improved transmission efficiency and reliability. Furthermore, this feature enables the optimization of resource allocation and prioritization of information transmission based on the evaluated relative information gain, thereby advantageously maximizing the expected information gain for the sink node(s).It is also possible that the procedure further comprises at least one of the following: - transmitting, by the central entity or cluster head, the aggregated information to one or more nodes of the distributed system; or - transmitting, by the central entity or cluster head, a statement about the available information to one or more nodes of the distributed system; or R. 415547 -. 7 -- Forwarding, by the central entity or cluster head, the information based on the evaluated or estimated relative information gain to one or more nodes of the distributed system, or - Requesting, by the central entity or cluster head, further information from one or more source nodes of the distributed system, where the further information is missing at the at least one sink node. It is possible that the procedure for transferring information in a distributed system includes further steps to enable the allocation of resources and the prioritization of information. These further steps enable more efficient and effective information transfer, taking into account the availability of information and its potential impact on the overall system.According to one embodiment, the relative information gain of the information available at one or more source nodes for the at least one sink node is estimated by the at least one sink node and / or by the one or more source nodes. That is, it is also possible for the at least one sink node to estimate the relative information gain of the information available to it. Preferably, the relative information gain is estimated by determining an information value for the information. Based on the estimated relative information gain, preferably based on the determined information value, an allocation of resources for transmitting the information can be requested. This has the advantage for the sink node that it can estimate the expected information gain from receiving specific data using a function or a model.This function could integrate factors such as the current state of knowledge at the sink node, the potential impact of new information on decision-making processes, and the reliability of the source nodes transmitting the data. It is also possible for the at least one sink node of the one or more sink nodes to estimate the relative information gain of the information available to that at least one sink node. R. 415547 -. 8 -Furthermore, it is possible that the function is based on an estimate of the divergence between probability distributions with respect to the information in order to determine the relative information gain of the information available at one or more source nodes. The divergence represents a measure of how a first probability distribution differs from a second probability distribution. It is possible that the described procedure for information transfer in a distributed system uses a function that takes into account an estimate of the divergence between probability distributions, thereby enabling the evaluation of the relative information gain and the subsequent allocation of resources to prioritize the transfer of specific information.This approach allows for a more precise evaluation of the available information at the sink nodes and the optimization of resource allocation to maximize the expected information gain. Considering the divergence between probability distributions can also facilitate a better understanding of the relationships and dependencies between different individual pieces of information within the distributed system. Furthermore, this enables a more precise assessment of how much new information a particular data element provides to the sink node compared to the existing estimated information. By focusing on the differences in the probability distributions, the function can effectively capture the uncertainty and novelty of the available information.It is also possible for the function to be based on a Kullback-Leibler divergence and / or an Angular similarity function and / or a Jensen-Shannon divergence and / or a population stability index function and / or a combination and / or supplement thereof. These metrics allow for a more nuanced assessment of how much new information a data element from source nodes contributes to the estimated information already available at the sink nodes. The choice of function depends on the specific characteristics of the distributed system and the nature of the information being transferred. R. 415547 -. 9 -Furthermore, the evaluation may also include: determining, using an evaluation function of the distributed system, the relative information gain, in particular an information value, as a function of the function. Using an evaluation function within the distributed system, it is possible to determine the relative information gain, in particular an information value. The evaluation function can analyze the relationship between the available information at the source nodes and the estimated information at the sink nodes. The function can incorporate probability distributions of the evaluated information to calculate the relative information gain precisely. This refined estimation process enables a more precise allocation of resources based on the maximized expected information gain for the nodes, in particular for a source node and / or a sink node.It is also possible for the function to assess the relative information gain that a local environmental model contributes to a global environmental model, based on the weighting of the precision and / or redundancy and / or dissimilarity of the respective information, in particular individual pieces of information, where the local environmental model comprises the information available at each of the one or more source nodes, and where the global environmental model comprises the information available at each of the one or more source nodes combined with information obtained from the network, where the one or more source nodes are also sink nodes. Therefore, the global environmental model includes local information from the local environmental model, i.e., information collected by the node itself and fused with information obtained from the network.By weighting these factors, the function can assess the potential benefits of integrating local observations into the broader understanding captured by the global environmental model. Furthermore, this feature allows for a more nuanced evaluation of the information value, recognizing that precision, redundancy, and / or dissimilarity are all important for constructing a comprehensive picture of the environment. The resulting R. 415547-. 10 -Prioritized information transfer can lead to improved awareness and, consequently, better-informed decisions at the sink nodes. Furthermore, the function is possible to be based on a weighted evaluation model that uses different metrics and / or scaling factors and / or user-defined functions based on covariance matrices and multidimensional expected values, for example, kinematic and posture state vectors, to determine the relative information gain. The scaling factors and / or weights are adjustable to prioritize precision over redundancy and / or dissimilarity of the available information. Distance metrics are based on a Bhattacharya distance, which measures similarity between probability distributions, and / or are based on a mean difference between LEM and GEM.and / or the user-defined functions are further based on a weighted sum of a determinant of the Kalman gain and the Bhattacharya distance, which specifies the weight of the mean difference and the covariance difference between the local and global environmental models. This allows the function, through adjustable weights and / or scaling factors, to prioritize either the precision and / or redundancy and / or dissimilarity of the available information. Furthermore, the function can be based on kinematic and posture state vectors, as well as covariance matrices, to determine the relative information gain, incorporating distance metrics such as the Bhattacharya distance to measure the similarity between probability distributions or the difference in means between the local and global environmental models. It is also possible for the procedure to further include: - Prioritizing,at the one or more source nodes, generating information depending on the evaluated relative information gain and depending on a defined information gain threshold, at the one or more source nodes, of at least one message based on the prioritization of the information, R. 415547 -, 11 -- At least one message is transmitted from one or more source nodes to at least one sink node. By prioritizing the information at the source nodes based on the evaluated relative information gain and a defined information gain threshold, the transmission of more relevant information can be ensured. This prioritization, in turn, enables the generation of messages at the source nodes that focus on the most valuable information. Consequently, transmitting these prioritized messages to the sink nodes maximizes the expected information gain for the system.It is further possible that: - the request for allocation, preferably the allocation of resources, is based on a defined information gain threshold; and / or - the request for allocation, preferably the allocation, is based on received information regarding the information gain threshold used by the other source node(s); and / or - the request for allocation, preferably the allocation, is based on an (estimated or determined) availability of resources in the network (specifically for the source node); and / or - the request for allocation, preferably the allocation, is based on a dynamically adjustable transmission rate depending on the evaluated relative information gain; - wherein the information gain threshold and / or the transmission rate are preferably transmitted to the at least one sink node, particularly as part of a message.The defined information gain threshold ensures that not every node always uses all available resources, but also transmits less if the information does not provide sufficient added value or information gain for the system. For example, if 20 stations detect the same truck on the highway, a 21st station that detects the truck from an even greater distance does not need to communicate this, regardless of whether it is already transmitting anything or not (i.e., whether it still has resources available). R. 415547 -. 12 -It is possible to improve the information transfer process in a distributed system by introducing an additional control layer. According to the invention, resource allocation can advantageously be based either on a defined information gain threshold or on a dynamically adjustable transfer rate that adapts to the evaluated relative information gain. This refinement enables more precise optimization of the information flow, as it allows for tailored decisions regarding resource allocation and transfer prioritization. The fixed information gain threshold provides a predictable and consistent approach, while the dynamically adjustable transfer rate offers greater flexibility and responsiveness to changing network conditions, ultimately optimizing the overall performance and efficiency of the distributed system.In a further aspect of the invention, a computer program, in particular a computer program product, can be provided, comprising instructions which, when the computer program is executed by at least one computer and / or evaluation function and / or data processing device and / or platform, cause the computer and / or evaluation function and / or data processing device and / or platform to execute the method according to the invention. Thus, the computer program according to the invention can have the same advantages as described in detail with reference to a method according to the invention. In a further aspect of the invention, a data processing device configured to execute the method according to the invention can be provided. For example, at least one computer can be provided as such a device, which executes the computer program according to the invention.The computer may include at least one processor that can be used to execute the computer program. It may also include non-volatile data storage in which the computer program can be stored and from which the computer program can be read by the processor for execution. R. 415547 -. 13 -Another aspect of the invention is an evaluation function for assessing the information gained from the data, comprising means for carrying out the method according to the invention. Thus, the evaluation function according to the invention can offer the same advantages as described in detail with reference to the method according to the invention. In another aspect of the invention, a platform can be provided, wherein the platform is configured as an infrastructure unit, e.g., a Road-Site Unit (RSU), or as a vehicle. In other words, the platform can be stationary or mobile. The vehicle can be a ground vehicle or an aircraft. The vehicle can be unmanned or manned.According to a further aspect of the invention, a computer-readable storage medium can be provided which comprises the computer program according to the invention and / or instructions which, when executed by at least one computer and / or evaluation function and / or data processing unit and / or platform, cause the computer and / or evaluation function and / or data processing unit and / or platform to execute the steps of the method according to the invention. The storage medium can be configured as a data storage device such as a hard disk and / or non-volatile memory and / or a memory card and / or a solid-state drive. The storage medium can, for example, be integrated into the computer.Another aspect of the invention is a message comprising a V2X message, wherein the V2X message includes information, the information being processed according to the steps of the inventive method. The message may further include the information gain threshold and / or the transmission rate. Additionally or alternatively, the message may include a measure or information representing the relative information gain of one, several, or all of the R. 415547 available at the one or more source nodes. 14 -The method represents or characterizes information and / or a distribution of the relative information gain. Thus, the message according to the invention disclosure can offer the same advantages as described in detail with reference to the method according to the invention. Furthermore, the method according to the invention can be implemented as a computer-implemented method. Alternatively or additionally, at least one of the disclosed method steps can be computer-implemented and / or automated. It is possible that the method according to the invention is used in a vehicle and / or infrastructure. The vehicle can, for example, be designed as a motor vehicle and / or passenger car and / or at least a partially automated / autonomous vehicle. The vehicle can have a vehicle device, e.g., for providing an autonomous driving function and / or a driver assistance system.The vehicle equipment can be designed to control, accelerate, brake, and / or steer the vehicle at least partially automatically. The infrastructure can be designed, for example, as a road element, traffic light, and / or traffic sign. Based on the method according to the invention, at least one control action for the vehicle can be initiated and / or executed. The control action can include at least one of the following: braking, steering, accelerating, overtaking maneuver, emergency braking, activation of an alarm system, activation of hazard warning lights, activation of turn signals, light control, or the like. For example, braking can also be initiated if the method according to the invention indicates that there are obstacles in the direction of travel and / or a collision is likely.It is also conceivable that, based on the classification, a lane and / or lane boundary is recognized in order to move the vehicle at least partially automatically within the lane as a result of the steering action. R. 415547 -. 15 -Further advantages, features, and details of the invention will become apparent from the following description, in which embodiments of the invention are described in detail with reference to the drawings. In this context, the features mentioned in the claims and in the description can be essential to the invention, either individually or in any combination. The figures show: Fig. 1: a method, a computer program, a storage medium, and a device according to embodiments of the invention; Fig. 2: a schematic diagram of a functional sequence according to embodiments of the invention; Fig. 3: another schematic diagram of message generation according to embodiments of the invention; Fig. 4: a schematic overview according to embodiments of the invention; and Fig. 5: a schematic view of a scenario according to embodiments of the invention.In the following figures, identical reference numerals are used for the same technical features even in different embodiments. Fig. 1 shows a method 100 for transmitting information in a distributed system. The method 100 comprises: In step 101, the available information at at least one sink node of the distributed system is estimated. In step 102, the relative information gain 30 of information 40 available at one or more source nodes of the distributed system is evaluated in relation to the estimated information 101 available at the at least one sink node, based on a function. The function depends on at least one probability distribution of the information available at the at least one sink node and / or at the one or more source nodes, for example, depending on covariances.In step 103, based on the evaluated relative information gain of 102, an allocation of resources in a network is requested to facilitate the transmission of R. 415547. 16 -to prioritize the information 40. This maximizes the expected information gain for the at least one sink node. Furthermore, Fig. 1 shows a node 10 or a device 10 for data processing, comprising a computer-readable storage medium 15. The medium 15 includes a computer program 50. Fig. 1 also shows a message 600, which includes a V2X message 610. The V2X message 610 includes information, preferably object data of objects 40. The message 600 can further include at least one other message type, which is processed by the method 100 according to the invention. Fig. 2 shows a schematic diagram of a functional process according to embodiments of the invention. In particular, Fig. 2 shows a schematic diagram for processing available information.Available information can include data from detected or perceived objects, or information such as temperature, humidity, position, speed, acceleration, or other measured or inferred data. Furthermore, Fig. 2 depicts two blocks 12, 20, where the first block 12 can be an evaluation unit 12 and the second block can be an evaluation function 20 or another evaluation unit 20. The evaluation function 12 can be designed as a pre-filter to pre-select or choose information such as object data for an object. The evaluation function 20 can, for example, further process available information based on a relative information gain, also referred to as information value, and decide, based on the available resources 203, whether the relevant information 20, 21 is transmitted or distributed 205.The evaluation unit 12 can be designed, for example, to select an object that has been detected or perceived by a sensor system of node 10. Therefore, the evaluation unit 12 can receive such information or data for further processing 202 or with regard to the generated message(s) 204 based on the available resources 203 201. R. 415547 -. 17 -In one embodiment, the further processing of the pre-selected information or data can be carried out, for example, by a further traffic relevance function, an evaluation function 20, or the like. In another embodiment, the information can be forwarded to another node or another function of a distributed system. In yet another embodiment, the evaluation unit 12 can receive information or data via a message from one or more nodes 201. The one or more nodes can be elements or components of a distributed system. Furthermore, Fig. 2 shows an evaluation function 20 that prioritizes pre-selected 202 information or object data. The (available) information or data or object data 202 can be selected for transmission 205 based on an estimated information value or information gain that they bring to the receiving sink node.The informational gain of available information, particularly its informational value, can be estimated depending on an information entropy model with respect to the evaluated information. The evaluation unit 12 illustrated in Fig. 2 can select the information based on the quality of object perception. For each perceived object, as illustrated, for example, in Fig. 3, a state vector and its covariance matrix, or specific elements thereof, should be available. Covariance is a measure of the measurement quality. This encompasses the information age in the sense that tracking filters, such as a Kalman filter or a particle filter, can lead to decreasing precision over time if only state predictions can be made without update steps.In other words, a lack of new measurements can increase the age of the information and lead to less precise state estimates that rely more on predictions based on transition models than on updated measurements. R. 415547 -. 18 -Thus, the evaluation unit 12 shown in Fig. 2 can use the covariance to estimate the object perception quality (OPQ) by defining precision or accuracy thresholds for the different parameters or components of the state vector. For example, a variance (or likewise the estimated absolute accuracy for a 95% confidence interval) of the x and y positions should be greater than or equal to a position accuracy threshold pa_thr, where pa_thr could be tied to a minimum accuracy required by the receivers to use the data (e.g., track accuracy * constant). Alternatively, a scalar value can be defined, e.g., based on the information entropy of the data. This scalar is then compared by the evaluation unit 12 to a threshold, instead of comparing each state vector component to a threshold.The pre-filtering and selection process by the evaluation unit 12 or the OPQ filter 12, before or after further filters such as a traffic relevance filter or a value information function 20, can be carried out in functional units as shown in Fig. 2 using the rules according to the invention. The function for object perception quality should have the following form. which means that the value decreases with increasing noise in % n' or remains constant, or any combination of the components of χ. n ·Σnxn the corresponding covariance matrix of its available components. Depending on the implementation, either a scalar or another vector (usually with n or fewer components) or a matrix (also usually with reduced dimensionality) can be output. Furthermore, the OPQ is then compared with threshold values ​​as part of the OPQ filter. These threshold values ​​can be different for each component. R. 415547 - 19 - be, for example, 2 m (two meters) for x, y, 0.3 m / s for v x , v y , or 4m (four meters) for z. In one embodiment, a weighted covariance matrix can be used to calculate f: where W = E (where E is the identity matrix) would give all components equal weight, thus restoring the original unweighted covariance matrix. In other embodiments, it is optionally possible to use a precision-based, a determinant-based, or an entropy-based approach: 1. Precision-based: Λweighted – Σweighted (positive finite) 2. Determinant-based: Σweighted (scalar ∈ [0, infinity], the lower the value, the higher the OPQ) 3. Entropy-based: H = ½ log ((2πe)nweighted 2 Σweighted) (scalar, ∈ [0, infinity], lower for higher OPQ). In further embodiments, the aforementioned threshold comparison can be performed, for example, for the precision-based, the determinant-based, or the entropy-based approach as follows: 1.Precision-based: diag(Λ) as a component-wise comparison with a weighted n-dimensional threshold vector, where rules could be as follows: each component must be greater than the comparison (since it is already weighted, this could also be expressed without loss of generality as diag(Λ) ≥ threshold for all n components). Another weighted embodiment could be that at least m of n components must be greater than (or equal to) the threshold; 2. Determinant-based: |Σ| ≤ threshold-weighted; 3. Entropy-based: H ≤ threshold-weighted. R. 415547 -. 20 -Only objects to which the above statements apply can pass the filter. The filter could also be defined conversely, and objects that do not pass would be filtered out. Fig. 3 illustrates another schematic diagram of message generation according to embodiments of the invention. Different generation rules can be used to create a message 300, as illustrated in Fig. 3. For example, the message 300 could be a collective perception message, a shared information message, or the like. Based on these generation rules, such a message 300 can include different elements or container types contained in each message 300. The message 300 can include a header element 304, which comprises a PDU header, an administration container, a station data container, and a generation delta time.The PDU header can include information about the protocol version or message type. The management container can include components relating to reference time and reference position. Furthermore, it can include a sensor information container 305, a free-space addendum 306, and a perceived object container 301. The sensor information container 305 can contain information about individual sensors or sensor systems from which the perceived object data originated. The kinematic and posture state of an object can be represented in a corresponding state vector, which includes different components relating to the kinematic and posture state. Additionally, a transmission frequency for message 300 can be set between 1 and 20 Hz, depending on the load of the communication channel.The detected objects 301 can be added to the message 300 based on various inclusion criteria, such as object quality, object type (e.g., VRU), or similar. Based on these criteria, objects 302 can be omitted due to a perceived object quality. The capacity of the communication channel can then limit how many of these objects are ultimately selected for transmission. Objects 303 that are more familiar to potential recipients are omitted because they are considered "redundant." The omitted R. 415547 -. 21 -Objects 303 can, however, be transmitted on other channels 320, provided they are available. If a Multi-Channel Operation (MCO) with different communication channels is available, parts of the information or object data can be offloaded to different communication channels 310, 320. This is illustrated in Fig. 3 by the two messages 300, 350, where message 350 can include elements 304, 303. The element or container 303 contains a portion of the information 320 to be offloaded. Furthermore, this message 350 also includes a header element 304. It is also possible that information, object data, or objects 303 may be omitted due to redundancy. Each communication channel can support a specific message size, which depends, among other things, on communication technologies such as 4G-V2X, ITS-G5, or 5G-V2X.Therefore, message segmentation may be necessary if too much information or object data is selected for transmission. In another embodiment (not shown), such as in distributed systems like a wireless sensor network, the sharing of information between different nodes in a network is often dynamic. Source nodes possess certain available information or data that is required by one or more sink nodes. The information or data can consist of numerous subsets of data, such as temperature, pressure, timestamp, position, velocity, etc., which is also referred to as the available information vector (AIV, which can be understood as a generalization of a system state vector). Furthermore, the sink nodes have certain information requirements, also referred to as the required information vector (RIV).Since this is a dynamic system, the size of both vectors, AIV and RIV, can vary over time and adapt to the conditions of the environment and the system. To exchange available information, the nodes of the (distributed) system are connected by wired or wireless R. 415547 -. 22 -Communication links are established. The bandwidths of these communication links can be limited depending on their physical properties, thus often presenting a complex optimization problem. Conventional systems employ different approaches, such as: periodic transmission of all available data to the source nodes (complete AIVs), but at a low frequency to limit the load on the communication nodes; transmission of parts of the AIVs at a higher frequency than others if these components are more critical to the system; or request-based transmissions, where sink nodes can request updates regarding specific components of their RIVs. While the first two approaches are rather static and do not take into account the information actually required at the sinks, the latter approach can consume some of the valuable communication resources for transmitting the requests.According to embodiments of the invention, information, data, or object data can be selected for transmission based on an estimated information value or information gain they bring to the receiving sink node. The process can include certain steps that can be performed at one or more source nodes. Based on a defined system structure, available information at one or more sink nodes of the distributed system can be evaluated. The value of receiving certain information or data, which, for example, can be a component of the AIV, can depend on the estimated prior knowledge of a respective receiving or sink node. This knowledge can be estimated in different ways. It is possible that the knowledge about the receiving or source node can be estimated (independently) without external information from one or more source nodes (R. 415547). 23 -This means that it is based solely on local information. Furthermore, it is possible that the transmitting source node can assume that it is aware of previously transmitted information. Depending on how quickly the network topology changes in a given case, only the most recently transmitted data should be assumed to be known. It is also possible that the transmitting source node can analyze the intrinsic entropy of each component of the AIV. For example, when communicating weather information, it can be assumed that the prevailing weather conditions are known to the receiving sink node(s). If the transmitting source node also receives information from other source nodes, it can assume that the information it receives from other source nodes has also been received by the sink node(s) and is therefore known to them.In another embodiment, a centralized entity or instance could exist that can receive and aggregate information or data transmitted from all source nodes. Optionally, the central instance can explicitly request missing information from the network. Furthermore, a hint of available information can be sent, for example, by using a Bloom filter or an ObjectIdList in the message, optionally enriching the message with additional information such as object positions. This central instance can then forward all aggregated information to the source nodes. As a further option, the central instance can forward information to the corresponding sink nodes based on the added value or relative information gain estimated by the central instance.This can be provided, for example, by using methods of relative information entropy. R. 415547 -. 24 -In another embodiment, a cluster head of a decentralized cluster-based network topology can aggregate information received from its so-called leaf nodes. Furthermore, the decision to forward information through the network can also be based on further steps described below. According to embodiments of the invention, a relative information gain of the available information, for example, an information value, can be estimated at one or more source nodes of the distributed system with respect to the evaluated information, depending on an information entropy model. The model specifies a relationship between two (probability) distributions of the respective information. After the sink nodes have been evaluated using one or a combination of the methods already described above (such as...),Given that information (centralized or decentralized) is available, the sink node can estimate the potential information gain of each of its (AIV) components for the one or more sink nodes. This can be estimated or calculated, for example, as follows: One option could be the use of an approach based on a Kullback-Leitler divergence (D. KL ) 2 based on D KL is a measure of the relative entropy of two probability distributions. It allows us to estimate the information gain provided by the component of the AIV relative to the knowledge of the receiving or sinking node. If N0 represents the AIV component of the source node and N1 the estimated knowledge of the sinking node about this property, Σ0 and Σ1 represent the covariance matrices of N0 and N1, and µ0 and µ1 are the means of N0 and N1, then D can KL as follows: A higher divergence then means a greater gain in information and thus a higher information gain or information value (VoI). R. 415547 - 25 - Another option is a different approach based on angular similarity (50). Cosine similarity is a measure of the similarity between two vectors and can be determined by calculating the cosine of the angle between them. For two components of the AIV, A and B, the cosine similarity is calculated as follows: When normalized, this is referred to as angular distance: The addition of the angular distance is called angular similarity Sθ. It is bound to the interval [0,1] and is calculated as follows: Further options can be based on other KLD derivatives such as the Population Stability Index (PSI). The PSI is a symmetric version of KLD and can be calculated as follows: PSl(P, Q) = DKL(P || F) DKL(Q || P), Or based on a so-called Jensen-Shannon Divergence (JSD). The JSD is a symmetric version of KLD and is bound to [0,1]. It can be calculated as follows: According to embodiments of the invention, resources in a network can be allocated based on the estimated (relative) information gain R. 415547 - 26 -Information is assigned to prioritize its transmission. After the information gain or value of each component of the AIV has been estimated, it must be evaluated which components should be transmitted and how often. This decision may require a reasonable compromise between providing high-quality information and avoiding overloading the transmission channel. To achieve this, if the communication channel itself and its current load are not monitored, one option is to use one or more fixed thresholds regarding information gain or value (defined a priori) to decide whether an AIV component should be included in a newly generated message at one or more source nodes. The message generation rate can be fixed.The other option may be based on the experience that a priori certain transmission patterns rarely exhibit optimal solutions to the aforementioned compromise. With constant monitoring of the channel, however, message generation can be dynamically optimized. A data rate available for the information dissemination or transmission service can be estimated, and this service can then be optimized based on the value of the information it transmits by regulating both the overall message generation frequency and the quantity and selection of available information or data that each message may contain. Fig. 4 presents a schematic overview according to embodiments of the invention. In particular, Fig.Figure 4 illustrates the influence of an information threshold value 430 or an information gain threshold value 430 on available information, such as multiple message objects 410, to be selected or prioritized for transmission. As shown in Figure 4, the threshold value 430 is specified such that the information or objects 41, 42, 43 may be transmitted starting from an information value 31, 32, 33 that is greater than the threshold value 430. Consequently, objects 44, 45, 46 are omitted or not considered for transmission because their R. 415547 -. 27 -The respective information values ​​34, 35, 36 are less than the specified threshold value 430. In another embodiment (not shown), an exemplary use case may be a wireless sensor network (WSN) comprising distributed sensor nodes for forest fire detection. A sensor node may include temperature and / or humidity sensors. Such a sensor node may assume that an information sink knows temperature and humidity values ​​if they are close to the usual values ​​for a given date and time. Whenever the sensor node takes a new measurement, it may assess the value of the new information, for example, using the angular similarity of the measurements and the expected values. For example, an expected measurement result may be a temperature T = 28 degrees Celsius, humidity H = 46%, which may be further processed as follows. The expected date and time are a temperature T = 30 degrees Celsius and a humidity H = 50%, which can also be formulated as follows. To determine a (relative) information gain, the node can calculate A value close to 1 means that the information is quite similar to the aforementioned expected value and should therefore only be transmitted if sufficient resources are available. R. 415547 - 28 - In the event of an unexpected measurement result, the temperature could be, for example, T = 71 degrees Celsius, and the humidity H = 5%, which is used as follows. The expected date and time are a temperature T = 30 degrees Celsius and a humidity H = 50%, which will be used as follows. To determine the (relative) information gain, the node can calculate The similarity evaluation used in this example is much smaller, which means a greater information gain for the recipient. Therefore, this measurement should be preferred when choosing which of the two measurements is to be transmitted. Fig. 5 illustrates a schematic view of a scenario according to embodiments of the invention. In particular, a traffic scenario is shown in which a first vehicle 61 and a second vehicle 62 detect a cyclist 70 or a bicycle 70. The cyclist can also be referred to as a vulnerable road user 70. The third vehicle 60, which wants to turn onto the main road on which the first vehicle 61 is traveling, cannot detect the cyclist 70 in this scenario illustrated in Fig. 5. This is because the view is blocked by a building 80. The third vehicle 60 can receive common messages from the other vehicles 61, 62. These messages can be R.415547 -. 29 -These are so-called Vehicle-to-Anything (V2X) messages. The third vehicle 60 can benefit from these messages transmitted by the other two vehicles 61 and 62. Due to the different perspectives, the second vehicle 62 can measure different information or data, such as the characteristics of the cyclist 70, with varying degrees of precision. For example, the first vehicle 61 can precisely measure the width and speed of the cyclist 70, but not the length; the reverse is true for the second vehicle 62. The first vehicle 61 can transmit its message first, and the second vehicle 62 can receive it. Furthermore, the second vehicle 62 can assume that the third vehicle 60 can also receive the same message from the first vehicle 61. In this embodiment shown in Fig. 5, the vehicles 60, 61, and 62 are participants in a distributed system for the shared use of information regarding the intersection depicted in Fig. 2.Such a system can also be called a Vehicle Ad Hoc Network (VANET). This type of network is set up for a certain period of time at the location "intersection," that is, as long as vehicles 60, 61, 62, and the cyclist are near this intersection. When the second vehicle, 62, can then generate a new message itself, it can estimate an information gain, such as the value of the information it can provide, based on a Kullback-Leibler divergence. The information value of a transmitted object can be determined from the relative entropy between the state of the object known to the VANET and the measured state. Since fusion algorithms like the Kalman filter usually assume normal distributions of the data, the relative entropy can be calculated from the respective example distributions; alternatively, Poisson, Rice, or other distributions can be used equally well. where r⃑LEM, r⃑VANET ∈ ℝk and Σ VANET is the corresponding covariance matrix of the information (or a subset thereof) as known to VANET. R. 415547 - 30 - LEM is understood as a local environmental model representing the vehicle's surroundings. GEM is understood as a global environmental model representing the overall situation, for example, a traffic scenario or a forest fire scenario. Since this is generally unknown, it can be calculated, for example, as follows: The relative entropy can now be obtained from the Kullback-Leibler divergence: By combining these equations and performing some algebraic calculations, one obtains the computationally less complex representation: Relative entropy is measured in Napier's digits. Division by ln 2 thus yields the divergence in bits. Therefore, D KL / ln 2 can be interpreted as the expected number of additional bits that must be transmitted by the node or vehicle to complete the network's knowledge of NLEM for a given NVANET. According to Gibbs' inequality, the relative entropy is always non-negative DKL ≥ 0, where equality holds only if NLEM = NVANET almost everywhere. If, instead, non-fused data is to be transmitted, the relative entropy can be obtained by replacing NLEM with NMeasurement and calculating DKL(NMeasurement ∥ NVANET). R. 415547 - 31 -In a next step, mapping to a different space may be useful in some cases. Preferably, this can be achieved by normalizing the information gain to the range [0, 1]. This could be accomplished, for example, by applying ƒ(DKL) = 1 – e−DKL, which maps DKL: [0, ∞] → [0, 1]. However, using DKL / ln (2) is an equally valid option (though then not normalized to [0, 1]). For each option (taking into account the fused or unfused data available at the transmitting stations), an analysis of the potential information gain for VANET can now be calculated. Various heuristics can be used for this purpose. A brute-force solution that yields an optimal result could, for example, examine all permissible combinations of data elements that can be enclosed in an object.Finally, given a certain amount of channel resources, the selection of objects and their level of detail can be modeled as a variant of the knapsack problem. Similar solutions could therefore be applied. In another embodiment, the shared information, which is transmitted or distributed, for example, via multiple V2X message types, can include trajectory data of trajectories. These trajectories can be described at varying levels of detail. The trajectories can be future-oriented (for maneuver planning and coordination) or past-oriented (for better correlation of measurements or traces, e.g., to ensure that an object A is still the same object A that was detected a few seconds ago). Currently, trajectories are usually scanned uniformly or based on simple rules, such as distances.In another embodiment, for example, vehicle X can travel straight ahead at a certain high constant speed, and another vehicle Y can just begin to drive a critical left turn. R. 415547 -. 32 -In this embodiment, the position of vehicle X would be subject to significantly greater changes. With a sampling threshold of, for example, 4 m, as in the case of CAM, this would result in a comparatively large number of trajectory points to be transmitted. For vehicle Y, the position changes much less, resulting in fewer trajectory points to be included (even considering further sampling thresholds, such as a course change of 0.5°). However, if the constant speed of vehicle X is known, most of the information about it is not particularly valuable, as the receivers can easily predict it. On the other hand, the left turn maneuver of vehicle Y can come as a surprise. However, taking this information gain into account would lead to a significantly more efficient allocation of resources.The greater the deviation from the expected results, the higher the number of resources allocated (in this case, the number of sampled trajectory points). For example, if vehicle X were to brake sharply, it would immediately be allocated more resources. However, for another vehicle Z, which is driving in a roundabout, fewer resources are needed, since its position, direction, speed, etc., can easily be estimated from the most recently transmitted trajectory points. As soon as Z leaves the roundabout, the value of the available information increases, ensuring that this information is transmitted as quickly as possible. In another embodiment, it can be assumed that, for example, two observations are available for a single vehicle.First, object-related observation of the environment, which may already have been communicated via V2X by other nodes, vehicles, or infrastructure elements such as a Road Site Unit (RSU). Second, self-observation or so-called ego-perception. R. 415547 -. 33 -These two observations may partially capture the same set of objects, but with different accuracy or precision, resolution, etc. (diverse observation), and partially capture different objects that contribute to expanding and / or completing the ego vehicle's field of view. Both the V2X data and the ego perception can differ. This means that available information, such as the position of a detected object (i.e., across multiple observations), may differ from the actual reality. The covariances may also differ. Using the over-the-air V2X data, the ego vehicle (e.g., as the source node) can decide whether its observation can improve the overall perception when it is communicated to other vehicles, Road Site Units (RSUs), or any node or ITS station.In one embodiment, V2X-based data and ego-vehicle observations can statistically represent object features with first- and second-order moments. The respective object features are encompassed in a state vector f, which is statistically characterized by the mean vector µ (first-order moment) and the covariance matrix Σ (second-order moment). Therefore, ego-vehicle data, also known as the local environment model (LEM), and V2X data are represented by (µ. LEM , Σ LEM ) and (µ V2X , Σ V2X) denoted. Both are available at the ego vehicle. The ego vehicle – as the source node – can prioritize its information based on its quality and the added value (information gain; information value) it can bring to other connected nodes compared to the data already available to them. An example of such priorities 30 is presented in Fig. 4, which illustrates how objects 40 are selected for transmission based on their information gain, such as their respective information values ​​31, 32, 33, 34, 35, 36 and a dynamic priority threshold 430. Furthermore, the ego vehicle can fuse the received V2X data into its local environment model (LEM) and obtains data with the moments (µ). GEM , Σ GEM V2X and the local data of the Ego vehicle are also referred to as the global environment model (GEM). R. 415547 - 34 - Findings from experimental cases show, for example: a) given µGEM = µ LEM = µ V2X and det(Σ LEM ) » det(Σ v2x ) ≈ det (Σ GEM ) -> Ego-vehicle observation is not helpful when fused with V2X data b) given µ GEM = µ LEM = µ V2X and det(Σ GEM ) ≈ det(Σ LEM ) << det (Σ v2x ) -> Ego-vehicle observation will be significantly helpful when fused with V2X data c) given µ GEM = µ LEM = µ V2X and det(Σ LEM ) ≈ det(Σ V2X ) < det (Σ GEM) -> Ego-vehicle observation is helpful when combined with V2X data d) given µLEM = µGEM - α, where α (real number) is a predefined threshold (can be derived based on association rules) -> Ego-vehicle observation can potentially be helpful when fused with V2X data, where cases a, b, and c are intuitive. Case d is based on the fact that the ego-vehicle makes different observations (assuming the criteria for accuracy and trustworthiness are met) than the shared perception based on the received messages. It is then more valuable to share this with other vehicles. The information value (VoI) at the ego-vehicle, as an example of a source node, can be defined as a metric to quantify the value of the ego-vehicle's observation in improving overall perception performance (i.e., the higher the VoI, the better if it is shared with others).Therefore, for case d: • VoI1 > VoI2 if |α1| >|α2|, where |α| < T / 2 can optionally be set (predefined threshold from the association function and its requirements). The following function characterizes the information gain or information value (VoI) for the source node information, such as the Ego vehicle data with feature vector f, described by (µLEM, ΣLEM). R. 415547 - 35 - Any function G that satisfies the specified conditions can be used as a VoI function. That is, any function G should be maximized, as for example, the norm ||µLEM – µGEM|| increases, and minimized as ||ΣLEM − ΣGEM|| or det(ΣLEM) – det(ΣGEM) decreases. That is, it should be monotonically increasing in ||µLEM – µGEM|| and monotonically decreasing in ||ΣLEM – ΣGEM|| or det(ΣLEM) – det(Σ GEMThe function G must be differentiable. To be sufficiently general, the norm and determinant can be replaced by any other functions with similar behavior. In other embodiments, some functions that satisfy the aforementioned rules and conditions are, for example, an information value based on the revised Kullback-Leibler divergence. The Kullback-Leibler divergence of two Gaussian distributions can be expressed as follows: (Alle: distribution of the fused data, Ev: distribution of the Ego vehicle data, K is the dimension of the feature vector) The formula above takes into account divergence in the mean rather than convergence. Therefore, the revised KL divergence reflects divergence in covariance separation and convergence in mean separation. Thus, the following applies: Therefore, the value of the information is inversely proportional to the divergence of the covariance and the convergence of the mean (revised KL divergence) and can be written as follows: VoI = A * exp (-B − RDKL + C). It should be noted that in the revised KL divergence, the -K term is crucial for normalization, dimensionality correction, and adjustment of the divergence measure. A, B, and C are constants. R. 415547 - 36 - It should also be noted that the only difference between D KL and RD KL The sign of the Mahalanobis distance term. In some cases, a reversal of -K to K may also be suitable for a better representation of the information value (VoI). Weighted sum of the determinant of the Kalman profit (for the merger) and the Bhattacharya distance. V oI = det( Σ −1 V2X( ΣV2X + ΣLEM)−1 + β(µLEM – µGEM)T ΣLEM (µL L E E M M –µGEM))where ΣV2X and ΣLEM are covariance matrices, μLEM and μGEM are mean vectors, and β is a custom scaling factor. Each user-defined function design can be based on dΣ = || Σ GEM – Σ LEM||F or Σ = det(ΣGEM – ΣLEM) or Σ = det( Σ GEM) – det( Σ LEM)dμ = || μ GEM – μLEM|| or Vo / = β1dμ – β2dΣ, where β is a custom scaling factor (weight). The value of the information function 20 can advantageously include both deterministic (e.g., determinant of covariance matrices) and probabilistic (e.g., Bhattacharya distance) measures. Users can define their own user-defined functions that include specific metrics for comparison. Examples include the Frobenius norm difference between covariance matrices or the Euclidean distance between means.The VoI function 20 can combine these different metrics using weighted totals, allowing users to advantageously prioritize certain factors based on their application requirements. R. 415547 -. 37 - The above explanation of the embodiments describes the present invention in the context of examples. Naturally, individual features of the embodiments can be combined with one another as desired, insofar as this is technically feasible, without departing from the scope of protection of the present invention.

Claims

R. 415547 - 38 -Claims 1. A method (100) for transmitting information in a distributed system, comprising: - estimating (101) available information at at least one sink node of the distributed system, - evaluating (102) a relative information gain (30) of information (40) available at one or more source nodes of the distributed system in relation to the estimated (101) information available at the at least one sink node based on a function, wherein the function depends on at least one probability distribution of the information available at the at least one sink node and / or at the one or more source nodes, and - requesting (103) an allocation of resources in a network based on the evaluated (102) relative information gain (30) to prioritize the transmission of the information (40) and thus maximize an expected information gain for the at least one sink node. 2.Method (100) according to claim 1, characterized in that - the distributed system is configured as a centralized system with a central entity, wherein the method further comprises receiving and aggregating the available information transmitted by one or more source nodes by the central entity, or - the distributed system is configured as a decentralized cluster-based network topology with a cluster head, wherein the method further comprises receiving and aggregating the available information transmitted by one or more source nodes by the cluster head. R. 415547 - 39 -3. Method (100) according to claim 2, characterized in that the method (100) further comprises at least one of the following: - transmitting, by the central entity or cluster head, the aggregated information to one or more nodes of the distributed system; or - transmitting, by the central entity or cluster head, information about the available information to one or more nodes of the distributed system; or - forwarding, by the central entity or cluster head, the information based on the evaluated (102) relative information gain (30) to the one or more nodes of the distributed system; or - requesting, by the central entity or cluster head, further information from the one or more source nodes of the distributed system, wherein the further information is missing at the at least one sink node. 4.

5. Method (100) according to one of the preceding claims, characterized in that the relative information gain of the information (40) available at the one or more source nodes for the at least one sink node is estimated by the at least one sink node and / or by the one or more source nodes.

6. Method (100) according to one of the preceding claims, characterized in that the function is based on an estimate of a divergence between probability distributions with respect to the information in order to determine the relative information gain of the information available at one or more source nodes. R. 415547 - 40 -the function is based on a Kullback-Leibler divergence and / or an Angular similarity function and / or a Jensen-Shannon divergence and / or a population stability index function and / or a combination and / or supplement thereof.

7. Method (100) according to one of the preceding claims, characterized in that the evaluation (102) further comprises: - Determining, on an evaluation function (20) of the distributed system, the relative information gain (30) as a function. 8.Method (100) according to any of the preceding claims, characterized in that the function assesses the relative information gain (30) that a local environment model (LEM) contributes to a global environment model, based on the weighting of the precision and / or redundancy and / or dissimilarity of the respective information, wherein the local environment model comprises the information available at each of the one or more source nodes and wherein the global environment model comprises the information available at each of the one or more source nodes, which is combined with information obtained from the network, wherein the one or more source nodes are simultaneously sink nodes.Method (100) according to one of the preceding claims, characterized in that the function is based on a weighted evaluation model that uses different metrics and / or scaling factors and / or user-defined functions based on covariance matrices and multidimensional expected values ​​to determine the relative information gain (30), wherein the scaling factors and / or weights are adjustable to prioritize precision over redundancy and / or dissimilarity of the available information, and / or. R. 415547 - 41 -- the distance metrics are based on a Bhattacharya distance, which measures a similarity between probability distributions, and / or on a mean difference between LEM and GEM, and / or - the user-defined functions are further based on a weighted sum of a determinant of the Kalman gain and the Bhattacharya distance, which indicates the weight of the mean difference (µ) and the covariance difference (Σ) between the local environment model and the global environment model.10.Method (100) according to one of the preceding claims, characterized in that the method (100) further comprises: -prioritizing the information at one or more source nodes depending on the evaluated (102) relative information gain (30) and depending on a defined information gain threshold (430), -generating at least one message (600) at one or more source nodes based on the prioritization of the information, -transmitting at least one message (600) at one or more source nodes to at least one sink node.Method (100) according to one of the preceding claims, characterized in that - the request (103) for the allocation of resources is based on a / the defined information gain threshold (430) and / or - the request (103) for the allocation of resources is based on received information regarding the information gain threshold used by the other source node(s) and / or - the request (103) for the allocation of resources is based on the availability of resources in the network, and / or - the request (103) for the allocation is based on a dynamically adjustable transmission rate depending on the evaluated (102) relative information gain (30). R. 415547 - 42 -- wherein the information gain threshold (430) and / or the transmission rate are preferably transmitted to the at least one sink node, in particular as part of a message (600).

12. Data processing device (10), comprising means for executing the method (100) according to any one of claims 1 to 11.

13. Evaluation function (20) for evaluating the information gain (30) of available information, comprising means for executing the method (100) according to any one of claims 1 to 11.

14. Platform configured as an infrastructure unit or as a vehicle, comprising the data processing device (10) according to claim 12 and / or the evaluation function (20) according to claim 13. 15.Computer program (50), comprising instructions which, when the computer program (50) is executed by at least one computer (10) and / or the data processing device (10) according to claim 12 and / or the evaluation function (20) according to claim 13 and / or the platform according to claim 14, cause the computer (10) and / or the data processing device (10) and / or the evaluation function (20) and / or the platform to execute the method (100) according to any one of claims 1 to 11.

16. Computer-readable storage medium (15) on which the computer program (50) according to claim 15 is stored.

17. Message (600), comprising a V2X message (610), wherein the V2X message (610) comprises information (40), characterized in that the information (40) is processed according to the steps of the method (100) according to any one of claims 1 to 11.

Images