Server system

Abstract

The invention relates to a server system. A server system is provided with a plurality of prediction servers and a mediation server that mediates the exchange of data between the plurality of prediction servers. One prediction server is provided with an acquisition means for acquiring a communication service log relating to one partial communication area. A prediction unit that predicts future communication service quality of the one partial communication area on the basis of the communication service log; and a communication unit that transmits a first prediction result obtained by the prediction unit to the mediation server and selectively receives a second prediction result obtained by another prediction server from the mediation server. The prediction unit uses the first prediction result and the second prediction result to predict the communication service quality between one node in the one partial communication area and the other node in the other partial communication area.

Description

Technical Field

[0001] This invention relates to the technical field of server systems for predicting communication quality. Background Technology

[0002] As such a system, for example, a system has been proposed that uses a radio wave propagation simulator to infer the wireless quality of the service area and predict the degradation of wireless quality at the future location of the wireless terminal (see Japanese Patent Application Publication No. 2009-278421). Summary of the Invention

[0003] The concept of a digital twin has been proposed. This digital twin uses data collected in the real world, which is then analyzed and simulated in a virtual space by AI (Artificial Intelligence), and fed back to the real world. In the field of communication services, realizing a digital twin requires, for example, the collection, accumulation, and analysis of massive communication service logs. Therefore, the overhead associated with data collection becomes a problem.

[0004] The present invention was made in view of the above-mentioned problems, and its objective is to provide a server system capable of suppressing the overhead associated with data collection.

[0005] One aspect of the present invention provides a server system comprising multiple prediction servers that predict the quality of service (QoS) of communication in a portion of a communication area each responsible for, and an intermediary server that mediates the exchange of data between the multiple prediction servers. One of the multiple prediction servers comprises: an acquisition unit that acquires communication service logs related to a portion of the communication area it is responsible for; a prediction unit that predicts the future QoS of the portion of the communication area, i.e., a first QoS, based on the communication service logs; and a communication unit that sends a first prediction result representing the first QoS to the intermediary server and selectively receives from the intermediary server a second prediction result predicted by other prediction servers among the multiple prediction servers. The prediction unit uses at least one of the first and second prediction results to predict the future QoS of a node in the portion of the communication area and other nodes in other portions of the communication areas responsible for by the other prediction servers, i.e., a second QoS. The server system further comprises an indication unit that, if the second QoS is lower than a predetermined quality, indicates a change in communication settings to the node. Attached Figure Description

[0006] The features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will now be described with reference to the accompanying drawings, in which the same reference numerals show the same elements, and wherein:

[0007] Figure 1 This is a conceptual diagram illustrating the concept of a server system according to an implementation method.

[0008] Figure 2 This is a block diagram illustrating the structure of a server system according to an implementation method. Detailed Implementation

[0009] Reference Figure 1 and Figure 2 This describes the implementation method of the server system.

[0010] exist Figure 1 In this system, server system 1 includes servers 10a, 10b, and 10c, and server 20. Servers 10a, 10b, and 10c are configured to communicate with server 20. For example, MQTT (Message Queue Telemetry Transport) can be used as the communication protocol between servers 10a, 10b, and 10c and server 20. That is, server system 1 can be a publish-subscribe type system. Furthermore, the number of servers in server system 1 is just an example and is not limited thereto.

[0011] Server 10a can obtain communication service logs from each node included in communication zone Ar1. Server 10a can use the obtained communication service logs to recreate the communication environment of communication zone Ar1 in virtual space. Server 10b can obtain communication service logs from each node included in communication zone Ar2. Server 10b can use the obtained communication service logs to recreate the communication environment of communication zone Ar2 in virtual space. Server 10c can obtain communication service logs from each node included in communication zone Ar3. Server 10c can use the obtained communication service logs to recreate the communication environment of communication zone Ar3 in virtual space. Furthermore, servers 10a, 10b, and 10c can be referred to as "digital twin servers".

[0012] For example, a node can be a vehicle with communication capabilities, a relay device, a server device, etc. Furthermore, a vehicle with communication capabilities can be a connected vehicle. For example, the communication service logs associated with a vehicle acting as a node can include at least one of the following: vehicle location, communication history, communication quality, and server response latency. Alternatively, the vehicle's location can also be the location of the communication device. Furthermore, communication areas Ar1, Ar2, and Ar3 can each be defined based on at least one of the cell covered by a wireless base station and the area governed by an aggregation station located upstream of each wireless base station.

[0013] Server 20 mediates the exchange of data between servers 10a, 10b, and 10c. For example, server 20 may send data obtained from server 10a to at least one of servers 10b and 10c. Server 20 may send data obtained from server 10b to at least one of servers 10a and 10c. Server 20 may send data obtained from server 10c to at least one of servers 10a and 10b. Furthermore, server 20 may be referred to as a "digital twin agent".

[0014] Next, refer to Figure 1 and Figure 2 The actions of server system 1 will be explained. Figure 2 In this configuration, server 10a includes a log processing unit 11, a log database 12, a communication environment prediction unit 13, a network setting formulation unit 14, and a communication device 15. Server 20 includes a prediction result distribution unit 21. Furthermore, servers 10b and 10c may have the same structure as server 10a. Vehicle 100, acting as a node, includes a log collection unit 110 and a network setting update unit 120.

[0015] The log collection unit 110 of vehicle 100 can send communication service logs to the server 10a, which is responsible for the communication area Ar1 in which vehicle 100 exists. The communication service logs related to vehicle 100 as a node may include, for example, historical location information of vehicle 100, the amount of communication data sent and received by vehicle 100, the throughput / latency at that time, the signal strength and S / N ratio (signal-to-noise ratio) received by vehicle 100 from the wireless base station, the server's data processing latency and load level, etc. Furthermore, the log collection unit 110 can periodically send the communication service logs to server 10a.

[0016] The log processing unit 11 of server 10a can obtain communication service logs sent from vehicle 100. The log processing unit 11 can further obtain communication service logs sent from nodes other than vehicle 100 included in communication area Ar1. The log processing unit 11 can perform prescribed processing or summarize the communication service logs obtained from multiple nodes included in communication area Ar1. The log processing unit 11 can store the summary results in log database 12. Furthermore, in addition to the latest summary results, the log database 12 can also store past summary results.

[0017] The communication environment prediction unit 13 of server 10a can predict, for example, the future behavior of vehicle 100 as a node and the communication performance between nodes within communication area Ar1 (e.g., between vehicle 100 and vehicle 101) based on the summary results of communication service logs stored in log database 12. That is, the communication environment prediction unit 13 can predict the future communication service quality of communication area Ar1. In this case, the communication environment prediction unit 13 can also use AI, which, when inputting the summary results of communication service logs, outputs the future behavior of vehicle 100 and the communication performance between nodes within communication area Ar1.

[0018] Furthermore, the future behavior of vehicle 100 can include its movement trajectory, communication volume, and communication timing. Additionally, the prediction algorithm described in the following literature can be used for predicting the position of vehicle 100 (e.g., its movement trajectory): Nachiket Deo, Mohan M. Trivedi; Convolutional Social Pooling for Vehicle Trajectory Prediction, in Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2018, pp. 1581-1589. Furthermore, the prediction of communication quality can utilize wireless propagation simulators, communication simulators, or regression analysis and machine learning.

[0019] The communication device 15 can send the prediction result, i.e., the first prediction result, from the communication environment prediction unit 13 to the server 20. The server 20 can store the first prediction result sent from the server 10a to a storage device (not shown). The prediction result distribution unit 21 of the server 20 can send the first prediction result to at least one of the servers 10b and 10c. For example, the topic related to the first prediction result can be associated with the first prediction result. The prediction result distribution unit 21 can send the first prediction result to the servers in servers 10b and 10c that have registered topics associated with the first prediction result based on the topics associated with the first prediction result. Furthermore, if there is no server that has registered topics associated with the first prediction result, the prediction result distribution unit 21 can discard the first prediction result.

[0020] Server 10b can predict the future quality of communication services in communication area Ar2 based on communication service logs (e.g., a summary of communication service logs). Server 10b can send a second prediction result representing the predicted quality of communication services to server 20. The prediction result distribution unit 21 of server 20 can send the second prediction result to at least one of servers 10a and 10c.

[0021] Server 10c can predict the future quality of communication services in communication area Ar3 based on communication service logs (e.g., a summary of communication service logs). Server 10c can send a third prediction result representing the predicted quality of communication services to server 20. The prediction result distribution unit 21 of server 20 can send the third prediction result to at least one of servers 10a and 10b.

[0022] For example, the communication environment prediction unit 13 of server 10a can also predict the relationship between vehicle 100, which is a node in communication area Ar1, and server 200, which is a node in communication area Ar2 (see reference). Figure 1 The future communication service quality between the two servers. In this case, the communication device 15 of server 10a can receive from server 20 the prediction results of server 10b responsible for communication area Ar2 (e.g., the second prediction result mentioned above). In addition, the communication device 15 can also receive from server 20, in addition to the prediction results of server 10b, prediction results of the communication service quality of more than one server responsible for each node on the communication path between vehicle 100 and server 200.

[0023] For example, the communication environment prediction unit 13 of server 10a may also use at least one of the future communication service quality of communication area Ar1 (e.g., the first prediction result described above) and the prediction result of server 10b (e.g., the second prediction result described above) to predict the future communication service quality between vehicle 100 and server 200. This communication service quality may include end-to-end latency. The communication environment prediction unit 13 may input a fourth prediction result representing the future communication service quality between vehicle 100 and server 200 to the network setting unit 14.

[0024] The network setting determination unit 14 can determine whether the communication service quality indicated by the fourth prediction result meets the required specified quality. If the communication service quality indicated by the fourth prediction result is lower than the specified quality, the network setting determination unit 14 can send a network setting instruction to the vehicle 100 via the communication device 15. Furthermore, if the communication service quality indicated by the fourth prediction result meets the specified quality, the network setting determination unit 14 may not output a network setting instruction. For example, the network setting instruction may include at least one of the following: updating the routing table, changing the communication destination application server, changing the communication unit, and changing the data transmission frequency, the detail of the transmitted data, and the compression rate.

[0025] For example, by updating the routing table, it is possible to forward communication traffic in a way that avoids congested network segments. For example, by changing the communication destination application server, it is possible to select a server with lower latency, envisioned as end-to-end. For example, changes to communication units may include at least one of switching between different cellular lines or using them concurrently, or at least one of switching between different types of communication networks (Wi-Fi / satellite communication, etc.).

[0026] The network setting update unit 120 of the vehicle 100, which is a node, can reflect the network setting instruction in the communication control upon receiving the network setting instruction.

[0027] Furthermore, whenever the communication environment prediction unit 13 predicts the future communication service quality of communication area Ar1, the server 10a can send the prediction result representing the newly predicted communication service quality to the server 20 via the communication device 15. When the server 10a sends the new prediction result to the server 20, the server 10a can also delete the past prediction results related to communication area Ar1 from the server 20.

[0028] For example, in the case that server system 1 is a publish-subscribe type system, server 10a can selectively receive from server 20 the prediction results that server 10a utilizes, among multiple prediction results indicating the quality of the communication service predicted by other servers (e.g., at least one of servers 10b and 10c). This reduces communication overhead.

[0029] (Technical effect)

[0030] In server system 1, servers 10a, 10b, and 10c collect communication service logs related to their respective communication areas Ar1, Ar2, and Ar3. For example, compared to a server system where only one server collects communication service logs, server system 1 in this embodiment can suppress the overhead associated with data collection.

[0031] In other words, to accurately predict communication service quality (QoS) parameters such as radio wave reception quality, network congestion levels, and server processing load, it is necessary to collect a massive amount of communication service logs from each node on the network without delay. In the server system 1 of this embodiment, multiple servers 10a, 10b, and 10c perform QoS prediction processing in a decentralized manner, thus narrowing the scope of each server's (10a, 10b, and 10c) QoS log collection. As a result, server system 1 can suppress the overhead associated with data collection.

[0032] The following describes the inventive solution derived from the embodiments described above.

[0033] One aspect of the invention provides a server system comprising multiple prediction servers that predict the quality of service (QoS) of communication in a portion of the communication area each is responsible for, and an intermediary server that mediates the exchange of data between the multiple prediction servers. One of the multiple prediction servers comprises: an acquisition unit that acquires communication service logs related to a portion of the communication area it is responsible for; a prediction unit that predicts the future QoS of the portion of the communication area, i.e., a first QoS, based on the communication service logs; and a communication unit that sends a first prediction result representing the first QoS to the intermediary server and selectively receives from the intermediary server a second prediction result predicted by other prediction servers among the multiple prediction servers. The prediction unit uses at least one of the first and second prediction results to predict the future QoS of a node in the portion of the communication area and other nodes in other portions of the communication areas responsible for by the other prediction servers, i.e., a second QoS. The server system further comprises an indication unit that, if the second QoS is lower than a predetermined quality, indicates a change in communication settings to the node.

[0034] In the above embodiments, "servers 10a, 10b and 10c" are equivalent to an example of "prediction server", "server 20" is equivalent to an example of "intermediary server", "log processing unit 11" is equivalent to an example of "acquisition unit", "communication environment prediction unit 13" is equivalent to an example of "prediction unit", "communication device 15" is equivalent to an example of "communication unit", and "network setting setting unit 14" is equivalent to an example of "instruction unit".

[0035] In one example of this server system, the node could be a vehicle with communication capabilities.

[0036] This invention is not limited to the embodiments described above. Appropriate modifications can be made without departing from the spirit or idea of ​​the invention as understood from the claims and description as a whole, and server systems with such modifications are also included within the technical scope of this invention.

Claims

1. A server system comprising: multiple prediction servers, each responsible for predicting the quality of service (QoS) of a specific communication area; and an intermediary server, which mediates the exchange of data between the multiple prediction servers. in, One of the multiple prediction servers has the following features: The acquisition unit acquires communication service logs related to a portion of the communication area that the prediction server is responsible for; The prediction unit, based on the communication service log, predicts the future communication service quality of the partial communication area, i.e., the first communication service quality. as well as The communication unit sends a first prediction result representing the first communication service quality to the intermediary server, and selectively receives a second prediction result predicted by other prediction servers among the plurality of prediction servers from the intermediary server. The prediction unit uses at least one of the first prediction result and the second prediction result to predict the future communication service quality, i.e., the second communication service quality, between a node in one partial communication area and other nodes in other partial communication areas handled by the other prediction server. The server system also includes an instruction unit that, when the quality of the second communication service is lower than the specified quality, instructs the node to change the communication settings.

2. The server system according to claim 1, wherein, The node is a vehicle with communication capabilities.

Images