Finite-State Machine for Mid-Call Feature Interaction Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing telecommunications systems face challenges in anticipating and resolving feature interactions during calls, particularly when multiple features are activated, leading to unexpected behavior or system failures, especially in distributed network environments.
Innovation Solution
The implementation of a finite-state machine and corresponding method for mid-call feature interaction detection and resolution, which determines interactions between activated features and manages their activation or deactivation based on predefined rules and priorities, ensuring that only one feature is active at a time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If design-time techniques are used to anticipate feature interactions, then feature interaction detection is performed before calls are set up, but it is difficult to anticipate interactions from third-party features added beyond the platform vendor's design
Solution Approach 1:
The system performs preliminary feature interaction detection during call setup by examining the call signature and comparing it against a database of known interaction patterns. This preliminary action identifies potential interactions before they manifest during the call, allowing the system to prepare appropriate resolution actions while maintaining adaptability to third-party features through pattern matching
Solution Approach 2:
The system implements runtime feedback mechanisms that monitor feature activations during calls and dynamically detect interactions as they occur. This feedback loop allows the system to adapt to third-party features by learning from actual runtime behavior and updating interaction detection accordingly
2Reliability
If runtime feature interaction detection techniques are used, then interactions are detected during calls, but detailed models are difficult to maintain in distributed networked environments and introduce calculation overhead infeasible to process during call setup
Solution Approach 1:
The system segments the feature interaction detection problem by dividing it into discrete call signature patterns and corresponding resolution actions. Instead of maintaining a single complex runtime model, the system uses a database of segmented interaction patterns that can be independently stored, maintained, and processed, reducing overall system complexity while improving maintainability
Solution Approach 2:
The system changes the parameter representation from detailed runtime models to simplified call signature patterns with associated resolution actions. This parameter transformation reduces the computational overhead from infeasible to manageable levels during call setup while maintaining sufficient accuracy for detecting feature interactions in distributed environments
3Adaptability or versatility
If multiple features are activated during a call, then feature functionality is enhanced, but unexpected or unwanted behavior occurs such as permitting policies to be circumvented or causing the call to fail
Solution Approach 1:
The system applies preliminary anti-action by proactively detecting potential feature interactions during call setup and preventing problematic combinations before they can cause call failures or policy circumvention. The system examines the call signature, identifies potential interactions against the pattern database, and blocks or adjusts feature activations that would create unstable conditions, thereby maintaining call reliability while allowing safe feature combinations
Data Source
AI summary
Techniques for detecting and resolving feature interactions during calls are disclosed. In particular, a finite-state machine and a corresponding method detect when a feature that is invoked during a call would interact with another previously-activated feature, and ensure that both features are not active simultaneously. Three different techniques for resolution are disclosed: in one technique, activation of the latter feature is always refused; in a second technique, the former feature is always deactivated and the latter feature is then activated; and in a third technique, one of the two features is selected to be the active feature—perhaps based on priorities assigned to the features—and the features are activated and/or deactivated accordingly.