IMS Call Screening Using IVR and ML to Block Robocalls
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing call screening methods for identifying and rejecting spam calls and robocalls in IMS-based telecommunications networks are limited, often reactive, and can result in increased call latency, poor call quality, and lack of regulatory controls, especially on lower-end devices.
Innovation Solution
Implementing network-based call screening using IP Multimedia Subsystem (IMS)-native network elements and machine learning (ML) analytics, including a large language model (LLM), to intercept and analyze inbound calls for context, tone, and purpose, and provide interactive voice response (IVR) options to subscribers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If device-based call screening is implemented, then call screening capability is provided, but call latency increases and call quality deteriorates
Solution Approach 1:
The patent introduces an IVR system as an intermediary between the caller and the subscriber. The IVR system handles the screening process by playing pre-recorded messages and collecting caller responses, thereby offloading the screening functionality from the end-device to the network infrastructure. This eliminates the need for device-based processing that causes latency and quality issues.
Solution Approach 2:
The patent replaces the mechanical processing approach (device-based real-time analysis) with a network-based automated system. The IVR system uses pre-configured screening logic and automated voice interactions to perform screening without requiring device processing power, thus avoiding the latency and quality degradation associated with device-based methods.
2Measurement precision
If non-standard call flows are used for call screening, then spam call identification is achieved, but regulatory controls are lost
Solution Approach 1:
The patent implements a standardized call flow that serves multiple functions: it screens for spam calls, maintains regulatory compliance, and works across different devices and network conditions. The IVR system uses standard telecommunication protocols and call flows that are universally accepted, allowing the same process to achieve both spam identification and regulatory compliance without requiring device-specific implementations.
3Measurement precision
If advanced device processing is used for call screening, then screening accuracy improves, but battery consumption increases and lower-end devices cannot function
Solution Approach 1:
The IVR system acts as an intermediary that performs all intensive processing tasks on the network side. The caller's device only needs to interact with the IVR system through standard voice commands or keypad inputs, while the actual screening analysis, spam detection, and decision-making are performed by the network-based IVR system. This eliminates the need for advanced device processing and associated battery consumption.
4Object-generated harmful factors
If reactive filtering methods are used, then some spam calls are blocked, but call latency increases and call quality suffers
Solution Approach 1:
The patent implements screening actions before the call is fully established or routed to the subscriber. The IVR system intercepts incoming calls and performs screening interactions (playing messages, collecting responses) during the call setup phase. Based on the caller's responses, the system can reject spam calls before they reach the subscriber, or prepare screening information in advance, thereby reducing latency and avoiding quality degradation that would occur with post-establishment filtering.
Data Source
AI summary
A system of a telecommunications network that uses IP Multimedia Subsystem (IMS) network elements to screen inbound calls. A Media Resource Function (MRF) intercepts an inbound call and uses various techniques to screen the call. The MRF redirects the call to an interactive voice response (IVR) system, which prompts the caller to state the caller's name, records the caller's response, and relays it to the called subscriber, who can then decide to accept or reject the call. The system can present a random number challenge to the caller. The system can use machine learning analytics and a large language model (LLM) to analyze the response's contents, accepting the call only if it relates to a topic allowed by the subscriber, perform audio analysis to determine whether the caller sounds robotic, or detect whether the response contains other known robocall markers.