Dynamic Network Selection Using Variable Kernel Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional handover algorithms for mobile communication devices are unable to adapt to dynamic user preferences and changing network conditions, relying solely on single attributes like signal strength and threshold-based policies, which limits their effectiveness in multi-network environments.
Innovation Solution
A method using variable kernel regression functions to calculate predicted utility values for multiple communication networks, considering various selection metrics such as availability, quality of service, and cost, and dynamically updating these values through a kernel learning process to determine optimal network switching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional threshold-based handover algorithms are used, then the handover decision process is simple, but the system cannot adapt to dynamic user preferences and changing network conditions
Solution Approach 1:
The patent implements dynamic adaptability by using kernel regression functions that continuously learn from observed data to update user preferences and network metrics. The system transitions from static threshold-based decisions to dynamic models that adapt to changing conditions, allowing the handover algorithm to evolve its behavior based on actual network performance and user preferences over time.
Solution Approach 2:
The kernel regression function operates autonomously by learning from observed data without requiring explicit reconfiguration. The system self-adjusts its parameters and preferences through continuous learning from network metrics and handover outcomes, reducing the need for manual intervention while maintaining adaptability to dynamic conditions.
2Measurement precision
If multiple selection metrics are considered for network selection, then the network selection accuracy is improved, but the computational complexity and data requirements increase
Solution Approach 1:
The patent combines multiple selection metrics (signal strength, network availability, cost, quality of service) into a unified kernel regression model. By merging these diverse metrics into a single learned function, the system achieves accurate network selection while managing computational complexity through a unified approach rather than separate analyses of each metric.
Solution Approach 2:
The system dynamically adjusts the importance and weighting of different selection metrics through the kernel regression learning process. Rather than treating all metrics equally or using fixed weights, the algorithm learns the optimal parameter configurations from observed data, allowing flexible adaptation to changing network conditions and user preferences.
3Reliability
If expert-defined cost functions are used for multi-criteria handover, then the handover decisions are systematic, but the system cannot adapt when preferences change
Solution Approach 1:
The kernel regression function implements continuous feedback learning by observing actual network performance and handover outcomes. The system uses this feedback to update its understanding of user preferences and network metrics, ensuring that handover decisions remain reliable while adapting to changing conditions over time through iterative learning.
Solution Approach 2:
The system performs preliminary learning and parameter estimation before making handover decisions. By pre-learning from observed data and establishing baseline preferences and metrics, the system ensures reliable decision-making is in place before actual handovers occur, while maintaining the capability to adapt as new data becomes available.
Data Source
AI summary
A method for determining whether to perform vertical handoff between multiple network. The method comprises obtaining a plurality of selection metrics for each network, calculating, for each of the other communication networks, a predicted utility value from at least the corresponding plurality of selection metrics using a variable kernel regression function, obtaining, for the current communication network, a second plurality of selection metrics; calculating a second predicted utility value for the current communication network from at least the corresponding second plurality of selection metrics using a second variable kernel regression function, comparing each of the predicted utility values for each of the plurality of other communication networks with the second predicted utility value and switching to one of the other communication networks having the highest predicted utility value, if the highest predicted utility value is greater than the second predicted utility value.