Set Top Box Channel Prediction via Neural Network Buffering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Interactive television networks face significant burst loads on distribution servers due to simultaneous user requests for channel changes, particularly at peak times, leading to inefficiencies in channel switching.
Innovation Solution
A system using a residential gateway with an adaptive AI system that predicts channel change requests by analyzing viewer behavior through a neural network-based channel viewing map, pre-loading likely channels and enabling instant channel changes by buffering content in advance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional channel change processing is used, then channel switching is performed on demand, but significant burst loads occur on distribution servers during peak times
Solution Approach 1:
The system performs preliminary actions by predicting which channels users are likely to request next and pre-loading those channel contents into buffer memory before the actual channel change request occurs. This eliminates the need for real-time content retrieval during channel changes, thereby reducing server load during peak times while maintaining fast channel switching performance.
2Loss of time
If channel contents are pre-loaded in advance, then channel change latency is reduced, but system complexity increases due to prediction algorithms
Solution Approach 1:
The prediction system uses feedback from historical channel viewing data to continuously improve its channel change predictions. By analyzing patterns in user behavior over time and adjusting prediction algorithms based on actual viewing habits, the system achieves accurate predictions without requiring overly complex infrastructure. The feedback mechanism allows the system to learn and adapt, reducing the need for sophisticated prediction models.
3Ease of operation
If all channels are made available instantly, then user satisfaction improves, but distribution server resources are overwhelmed
Solution Approach 1:
Instead of making all channels available to all users simultaneously, the system applies local quality by pre-loading only the specific channels that individual users are likely to request, based on their viewing history and current context. This targeted approach ensures that each user experiences fast channel changes for their relevant channels while conserving server resources that would be wasted on pre-loading unrelated channels for other users.
Data Source
AI summary
Aspects of the subject disclosure may include, for example, a method comprising obtaining channel change data for a media processor coupled to a gateway; the channel change data relates to channel changes requested during recurring time periods. A channel viewing profile is determined for the media processor, and a channel viewing map is generated corresponding to a portion of the channel viewing profile for one of the time periods; the channel viewing map comprises a list of predicted channels likely to be requested during a recurrence of the time period. Signals are provided to the media processor corresponding to the predicted channels, each of the predicted channels thereby being available for an instant channel change responsive to a channel change request. Other embodiments are disclosed.