Distributed Network Framework for Joint Computation and Caching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies lack an optimized framework for joint computation, caching, and request forwarding in data-centric computing-based networks, which is essential for achieving high throughput and low latency in distributed computing environments.
Innovation Solution
The proposed framework provides a distributed and adaptive method for joint computation, caching, and request forwarding by using computation request counters and data request counters to determine optimal policies for each network node, enabling superior performance in request satisfaction delay across various network topologies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If centralized cloud architecture is used, then cost efficiency and scalability are improved, but request satisfaction delay increases for delay-sensitive applications
Solution Approach 1:
The patent segments the centralized cloud architecture into a hierarchical distributed structure with core network nodes, edge nodes, and user equipment. This segmentation allows computation and data storage to be distributed at multiple levels, enabling delay-sensitive applications to access data closer to the user while maintaining overall system scalability and cost efficiency through shared resources.
Solution Approach 2:
The patent introduces a new dimensional aspect by implementing in-network computation and caching capabilities at intermediate network nodes. This adds a computational dimension to traditional data transmission, allowing data to be processed and stored within the network itself rather than only at endpoints or centralized clouds, thereby reducing latency without sacrificing scalability.
2Loss of time
If dispersed computing paradigm is adopted, then request satisfaction delay is improved, but implementation complexity increases
Solution Approach 1:
The patent implements a universal framework where network nodes can perform multiple functions including data forwarding, computation execution, and data caching. This multi-functionality reduces implementation complexity by consolidating operations that would otherwise require separate specialized components, while maintaining low latency through coordinated distributed operation.
Solution Approach 2:
The patent incorporates feedback mechanisms where nodes exchange computation request counters and data request counters to dynamically adjust their behavior. This feedback loop enables adaptive decision-making about which nodes should execute computations and which should forward data, simplifying the control logic compared to static distributed systems while maintaining optimal performance.
3Productivity
If data-centric computation is performed, then throughput is improved, but resource utilization optimization becomes more difficult
Solution Approach 1:
The patent implements preliminary action through in-network caching, where data objects are stored at intermediate nodes before final consumption. This pre-positioning of data enables high-throughput computation by eliminating repeated data retrieval operations, while the distributed caching mechanism automatically optimizes resource utilization across the network without requiring centralized coordination.
Solution Approach 2:
The patent dynamically changes operational parameters at different network nodes based on local conditions. Nodes adjust their behavior by modifying computation request counters and data request counters, allowing the system to adapt resource allocation to actual demand patterns. This parameter-based control simplifies optimization compared to fixed resource allocation while maintaining high throughput for data-centric workloads.
Data Source
AI summary
A framework for joint computation, caching, and request forwarding in data-centric computing-based networks comprises a virtual control plane, which operates on request counters for computations and data, and an actual plane, which handles computation requests, data requests, data objects and computation results in the physical network. A throughput optimal policy, implemented in the virtual plane, provides a basis for adaptive and distributed computation, caching, and request forwarding in the actual plane. The framework provides superior performance in terms of request satisfaction delay as compared with several baseline policies over multiple network topologies.