Communication Speed Prediction Model Update Frequency Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing communication speed prediction apparatuses, such as those using neural networks, face challenges in properly updating their prediction models to accurately forecast communication speeds between moving communication apparatuses, like drones and ground base stations, due to variations in relative positional relationships.
Innovation Solution
A communication speed prediction apparatus that includes a position prediction unit, a speed prediction unit, an extraction unit, a training unit, and a model update unit, which predicts future communication speeds by using a prediction model updated based on relative positional relationships, trained with past position and speed information, and adjusts the update frequency based on the states of the apparatuses and communication environment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the prediction model is continuously updated using all available training information, then the prediction accuracy is improved, but the processing load and time consumption increase
Solution Approach 1:
The patent extracts only the necessary training information required for model updates rather than processing all available data. The extraction unit selectively identifies and extracts training samples from communication history information, reducing the volume of data that needs to be processed while maintaining prediction accuracy.
Solution Approach 2:
The patent applies partial action by updating the prediction model only when necessary rather than continuously. The model update is triggered based on specific conditions such as when extraction results indicate improved prediction accuracy is needed, avoiding unnecessary processing cycles and reducing overall processing load.
2Measurement precision
If the prediction model is updated frequently, then the prediction accuracy is improved, but the time consumption and processing overhead increase
Solution Approach 1:
The patent implements periodic action by updating the prediction model at specific intervals or under specific conditions rather than continuously. The model update occurs periodically when the extraction unit identifies that training information is available and update conditions are met, balancing accuracy improvement with time efficiency.
Solution Approach 2:
The patent uses feedback mechanisms where the extraction unit evaluates whether updating the prediction model will actually improve prediction accuracy. Based on this feedback assessment, the system decides whether to proceed with the update, avoiding unnecessary updates that would consume time without providing benefits.
3Measurement precision
If more training information is used for model training, then the prediction accuracy is improved, but the device complexity and resource requirements increase
Solution Approach 1:
The extraction unit extracts only the essential training information from the available communication history data, filtering out redundant or less useful samples. This selective extraction reduces the complexity of the training process while maintaining the quality and accuracy of the prediction model.
Solution Approach 2:
The patent uses temporary or disposable training samples that are extracted from communication history, processed for training purposes, and then discarded or replaced. This approach allows the system to use diverse training data without permanently storing or managing large complex datasets, reducing device complexity.
Data Source
AI summary
A communication speed prediction apparatus (1) includes: a position prediction unit (111) for predicting a future positional relationship between first and second communication apparatuses (B, D); a speed prediction unit (112) for predicting the future communication speed between the first and second communication apparatuses by using a predicted result by the position prediction unit and a prediction model (PD) for predicting the future communication speed between the first and second communication apparatuses; an extraction unit (113) for extracting a training information from an information source (121) including information relating to a past positional relationship between the first and second communication apparatuses and information relating to a past communication speed between the first and second communication apparatuses; a training unit for training a parameter of the prediction model by using the training information; and a model update unit (115) for updating, by using the trained parameter, the prediction model used by the speed prediction unit, the model update unit changes, based on at least one of a state of the communication speed prediction apparatus and a state of the first and second communication apparatuses, an update frequency by which the prediction model is updated.