Master's Thesis

FediRouter: Efficient Server Selection in Federated Networks

Miguel Matos Saldanha Albuquerque — 2026

Key information

Authors:

Miguel Matos Saldanha Albuquerque (Miguel Matos Saldanha Albuquerque)

Supervisors:

Luís Alexandre Duque Moreira de Sousa (Luís Alexandre Duque Moreira de Sousa); Flávio Nuno Fernandes Martins (Flávio Martins)

Published in

May 25, 2026

Abstract

Classical Information Retrieval systems rely on globally consistent and centrally accessible indexes to support retrieval and recommendation tasks. Federated social networks, however, deliberately abandon these assumptions by distributing data across independently operated servers, each maintaining its own storage, indexing structures, and policies. This architecture introduces major challenges for network-wide retrieval, since no single node maintains a complete view of the network. This work proposes FediRouter, a lightweight federated resource selection approach capable of recommending relevant servers without requiring synchronous inter-server communication or complete knowledge of peer-local data. The proposed solution models server selection as a probabilistic retrieval task and investigates the use of the Taily algorithm to estimate server relevance using only compact locally available statistical summaries. The effectiveness of the approach is evaluated against voting-based ranking methods, namely Borda Count, Condorcet, and Reciprocal Rank Fusion. Experimental results, assessed through Precision, Precision@k, Average Precision@k, and Mean Average Precision, show that the Taily-based approach consistently outperforms the baseline methods. Additionally, the efficiency and scalability of the system are analysed through various experiments that exhibit good latency and throughput, demonstrating the feasibility of deploying a lightweight federated server recommendation service in real environments.

Publication details

Authors in the community:

Supervisors of this institution:

Fields of Science and Technology (FOS)

electrical-engineering-electronic-engineering-information-engineering - Electrical engineering, electronic engineering, information engineering

Publication language (ISO code)

eng - English

Rights type:

Embargoed access

Date available:

May 1, 2027

Institution name

Instituto Superior Técnico