Dynamic Event Caching for Network Traffic Trends
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current caching methods, such as LRU and top-value schemes, are not sufficiently perceptive to short-term trends in computer network traffic, leading to inefficiencies in caching search results and increased latency in search engines.
Innovation Solution
A method using time-discounting mathematical formulas and priority queues to dynamically identify and cache currently dominant events in computer network traffic, ensuring that recent events are prioritized over past ones, allowing for real-time adaptation to changing traffic patterns and efficient use of cache resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If LRU or top-value caching methods are used, then cache space is utilized, but short-term trends are not detected and latency increases
Solution Approach 1:
The patent implements dynamic caching by continuously monitoring event rates and automatically adjusting which events are cached based on real-time trend detection. The system transitions from static LRU or top-value methods to a dynamic approach where cache eligibility is determined by detected short-term trends, allowing the cache to adapt to changing traffic patterns and reduce latency for emerging events.
Solution Approach 2:
The system incorporates feedback mechanisms by monitoring event rates and using this information to detect short-term trends. The detected trends then feed back into the caching decision process, creating a closed-loop system that continuously optimizes cache content based on observed traffic patterns, thereby reducing latency for trending events.
2Loss of time
If all search results are cached, then access time is reduced, but computer resources are excessively consumed
Solution Approach 1:
The patent applies local quality by differentiating between events based on their trend characteristics. Instead of uniformly caching all events or using a single caching policy, the system identifies events with short-term trends and applies caching selectively to these specific events. This localized approach ensures cache resources are concentrated on events that will benefit most from caching, reducing overall resource consumption while maintaining low access times for trending events.
Solution Approach 2:
The system performs partial caching by selecting only a subset of events for caching based on detected short-term trends, rather than caching all events. This partial action approach optimizes the balance between cache resource usage and access time reduction, caching enough events to significantly improve performance without consuming excessive computer resources.
3Productivity
If caching is applied to all events, then reuse is maximized, but cache efficiency for trending events decreases
Solution Approach 1:
The patent enhances cache efficiency by applying caching selectively to events with detected short-term trends rather than uniformly to all events. This localized quality approach ensures that cache resources are allocated to events with the highest potential for reuse, maximizing cache efficiency for trending events while using fewer overall cache resources.
Solution Approach 2:
The system changes the parameter of cache eligibility from static (LRU position or predetermined top values) to dynamic, based on detected short-term trends. By monitoring event rates and identifying trending events, the system adjusts cache content to match actual usage patterns, thereby improving cache efficiency and ensuring that cached events are more likely to be reused.
Data Source
AI summary
Methods and apparatus, including computer program products, implementing and using techniques for identifying currently dominant events in a computer system. Live computer network traffic including current events is monitored on the computer network. A subset of monitored events is selected. Event records are created for the subset of events. The event records include timestamps indicating when the event record was adjusted and counts indicating the popularity of an event when the event record was adjusted. Older events have lower counts relative to newer events. A first set of event records identifying currently dominant events is stored in a first priority queue on a computer and is cached in a cache on the computer. A second set of event records identifying currently not dominant events is stored in a second priority queue on the computer and is not cached.