Ad-hoc Node Message Period Adaptation via Neural Network
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication networks of ad-hoc nodes face inefficiencies in periodic message transmission, leading to excessive bandwidth consumption when the network is stable and insufficient discovery of neighborhood changes when it is unstable, due to fixed sending periods that do not adapt to network conditions.
Innovation Solution
A method using a neural network to dynamically adapt the sending period of periodic messages based on the stability of neighborhood links, calculated from parameters such as geographical position, signal strength, and connection/disconnection history, allowing for real-time optimization of message frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a low fixed sending period is used, then neighborhood discovery is efficient, but bandwidth consumption becomes excessive
Solution Approach 1:
The patent applies dynamics by transitioning from a fixed sending period to a dynamic sending period that adapts to network conditions. The system continuously monitors neighborhood stability parameters and adjusts the message sending period in real-time, allowing it to be short when topology changes frequently and long when the network is stable, thus resolving the contradiction between discovery efficiency and bandwidth consumption.
Solution Approach 2:
The patent changes the parameter of message sending period from a constant value to a variable that depends on network conditions. By introducing parameters such as neighborhood stability, link quality, and topology change detection, the system dynamically adjusts the sending period to optimize both discovery efficiency and bandwidth usage under different operational scenarios.
2Loss of energy
If a high fixed sending period is used, then bandwidth consumption is reduced, but neighborhood discovery becomes insufficient
Solution Approach 1:
The system dynamically adjusts the sending period based on detected network conditions. When topology changes are detected or neighborhood stability decreases, the sending period is automatically reduced to ensure adequate discovery. This dynamic adaptation allows the system to maintain efficient discovery only when necessary, rather than using a consistently high frequency.
Solution Approach 2:
The patent implements feedback mechanisms where nodes monitor neighborhood stability, link quality, and topology changes, then use this information to adjust the message sending period. This closed-loop control ensures that bandwidth is consumed at high rates only when the network conditions warrant intensive discovery, automatically reducing transmission frequency when the network is stable.
3Adaptability or versatility
If manual configuration of decision thresholds is used, then adaptability to specific network families is achieved, but system complexity and configuration effort increase
Solution Approach 1:
The patent enables the system to self-configure by automatically learning optimal decision thresholds through machine learning algorithms. Instead of requiring manual configuration for each network family, the system observes network behavior patterns and autonomously determines appropriate thresholds for adapting the message sending period, thereby maintaining adaptability while eliminating configuration complexity.
Solution Approach 2:
The patent replaces the mechanical/manual configuration system with an intelligent system based on machine learning and automated decision-making. The neural network or learning algorithm automatically determines optimal parameters and thresholds based on observed network conditions, substituting human configuration efforts with autonomous system intelligence that adapts to different network families without manual intervention.
Data Source
Figure 1~2
Figure 3
AI summary
The invention relates to a method for dynamically adapting the sending period, T_diff, of periodic messages emitted by an ad-hoc node (N) comprising a neural network (24), further comprising the following steps implemented by said node every T_calc seconds: - provision of an updated value of the parameter vector(s) qualifying the current stability of the neighbor links of the node; - calculation by the neural network (24) of scores (Q_value_1,...., Q_value_n) each associated with a distinct value of sending period (T1, ..., Tn) as a function of said updated value of the parameter vector(s), said score evaluating the relevance of implementing said associated value of sending period as a function of said parameter vector(s); - selection of one of said distinct candidate values (T1, ..., Tn) of sending period as a function of the calculated scores; - assignment of the selected value to T_diff for the duration T_calc.