Predictive Content Delivery for Ad Request Congestion Relief
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Content providers face significant processing delays and network congestion due to simultaneous requests for advertising content from playback devices during high-viewership events, leading to poor viewing experiences.
Innovation Solution
Implementing a system that predicts events within a content stream and inserts predicted markers, allowing playback devices to wait a randomly generated time before requesting advertising content, thereby reducing simultaneous requests and alleviating network congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the system processes advertising content requests in real-time during high-viewership events, then the viewing experience is degraded due to processing delays, but if the system caches content in advance, then network congestion is reduced
Solution Approach 1:
The system performs preliminary actions by predicting future advertising content requests based on scheduled events and popular content patterns. Before the actual viewing event occurs, the system pre-loads and caches the predicted advertising content at edge servers, so that when millions of users simultaneously request content during high-viewership events, the advertising content is already available locally, eliminating processing delays and ensuring reliable content delivery.
2Productivity
If the system allows simultaneous requests from millions of playback devices, then network congestion occurs, but if requests are delayed, then content delivery timing is compromised
Solution Approach 1:
The system introduces edge servers as intermediary components between the central content delivery network and millions of playback devices. These edge servers act as local caches that receive and store predicted advertising content in advance. When simultaneous requests occur during high-viewership events, the edge servers fulfill requests locally without forwarding all requests to the central system, thereby eliminating network congestion while maintaining fast content delivery speed.
3Device complexity
If the system predicts and caches advertising content in advance, then instantaneous load on the decision system is reduced, but additional processing overhead is introduced for prediction and caching
Solution Approach 1:
The system performs prediction and caching actions in advance during low-demand periods, when the decision system is not under instantaneous load. By analyzing historical data, scheduled events, and content popularity patterns before peak viewing times, the system pre-determines which advertising content will be requested and caches it at edge servers. This shifts the processing workload from high-demand periods to low-demand periods, reducing decision system load during events while maintaining overall processing efficiency.
Data Source
AI summary
Methods and systems are described herein to predict events within a content stream. Based on the predicted events, information may be embedded within the stream. A computing device that receives the content stream may detect the predicted event and may wait a randomly generated time before requesting information associated with the predicted event, such as requesting an advertisement to output. After detecting the predicted event, the computing device may wait a randomly generated time before requesting an advertisement. The system may respond to the computing device with an instruction or with supplemental content, such as with an advertisement for later output. The computing device may cache the advertisement. The computing device may detect the actual marker indicating a timestamp for when the advertisement should be output. The computing device may output the advertisement at the time indicated by the timestamp.