We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.IT

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Theory

Title: Collaborative Artificial Intelligence (AI) for User-Cell association in Ultra-Dense Cellular Systems

Abstract: In this paper, the problem of cell association between small base stations (SBSs) and users in dense wireless networks is studied using artificial intelligence (AI) techniques. The problem is formulated as a mean-field game in which the users' goal is to maximize their data rate by exploiting local data and the data available at neighboring users via an imitation process. Such a collaborative learning process prevents the users from exchanging their data directly via the cellular network's limited backhaul links and, thus, allows them to improve their cell association policy collaboratively with minimum computing. To solve this problem, a neural Q-learning learning algorithm is proposed that enables the users to predict their reward function using a neural network whose input is the SBSs selected by neighboring users and the local data of the considered user. Simulation results show that the proposed imitation-based mechanism for cell association converges faster to the optimal solution, compared with conventional cell association mechanisms without imitation.
Subjects: Information Theory (cs.IT)
DOI: 10.1109/ICCW.2018.8403664
Cite as: arXiv:1809.01254 [cs.IT]
  (or arXiv:1809.01254v1 [cs.IT] for this version)

Submission history

From: Ali Taleb Zadeh Kasgari [view email]
[v1] Tue, 4 Sep 2018 21:52:06 GMT (1929kb,D)

Link back to: arXiv, form interface, contact.