Computer Science

基于熵的结构可观测性框架

Entropic Structural Observability Framework

A framework for routing-independent post-simulation analysis of Network-on-Chip (NoC) traces. It converts packet-level events into probability distributions over resources (routers, links, time windows) to generate structural signatures including normalized entropy, mutual information, and topology-aware spatial statistics. / 一种用于片上网络（NoC）轨迹后仿真分析的框架。它将包级事件转换为资源（路由器、链路、时间窗口）上的概率分布，以生成包含归一化熵、互信息和拓扑感知空间统计的结构签名。

Knowledge type: 工艺技术
First proposed or discovered: 2026
Proposer: Ahmed Mesellem, Mohammed Mana
Creator: 农场主
Last updated: 2026-11-14 18:06:49

Sub-knowledge points

Typed packet-level event conversionProbability distribution generation over resourcesStructural signature constructionDiagnostic layer operation

Higher-order knowledge points

Network-on-Chip (NoC) evaluationSimulation trace analysisRouting policy evaluation

Main parameters and performance indicators

Normalized entropy: Measures the uncertainty or randomness of the resource usage distribution.Effective support: The range of resources actually utilized in the trace.Concentration: The degree to which usage is focused on specific resources.Spatiotemporal mutual information: Quantifies the dependency between spatial and temporal resource usage.Distribution drift: The change in the resource usage distribution over time.Topology-aware spatial statistics: Resource usage patterns relative to the network topology.

Open problems

No unresolved problem recorded yet

Related literature

ranked by community stars

A Trace-Based Structural Observability Framework for Network-on-Chip Routing EvaluationOpen source

基于熵的结构可观测性框架

Sub-knowledge points

Higher-order knowledge points

Main parameters and performance indicators

Open problems

Related products and links