Description
TL;DR: A mathematical study of the bias-variance tradeoff that appears when distributed nodes repeatedly average data over unreliable, noisy links.
About This Document
Soummya Kar and José M. F. Moura present two approaches. The A-ND algorithm uses slowly decaying weights to obtain asymptotically unbiased consensus with low variance but slower convergence. The A-NC approach runs fixed-weight consensus for a limited number of iterations, restarts it, and averages results across repeated runs. The paper analyzes convergence, mean-square error, run length, noise, and random topology under general communication assumptions.
Why Bitcoin People May Care
This predates Bitcoin and does not select a transaction ledger under adversarial mining. It belongs in the distributed-systems background because it shows how even honest nodes face competing goals when communication is imperfect. Consensus quality, convergence speed, and noise do not collapse into one number simply because every participant runs the same algorithm.
What You Receive
Purchase reveals the verified preserved PDF location for this document.
Document Details
- Title: Distributed Consensus in Sensor Networks with Link Failures and Channel Noise
- Author / organization: Soummya Kar and José M. F. Moura
- Year: 2008
- Language: English
- Document type: Research paper
- Pages: 30
CacheRat provides researched source-location and document-identification information. CacheRat does not claim ownership of the underlying third-party documentation.
