Distributed-Optimization-with-Centralized-Refining

Citation Author(s):
Jiyang
Bai
Xianbin
Wang
Submitted by:
Jiyang Bai
Last updated:
Tue, 03/19/2024 - 16:30
DOI:
10.21227/kzja-gd50
Data Format:
License:
0
0 ratings - Please login to submit your rating.

Abstract 

Distributed-Optimization with Centralized-Refining (DO-CR) mechanism to achieve more efficient resource allocation by engaging both access point and all devices. Specifically, the new DO-CR mechanism first utilizes the distributed processing capacity of all devices, allowing them to optimize their own resource allocation schemes through a new resource reservation and reporting technique. Then a centralized optimizer generates a graph of resource trading topology based on individual optimization results and achieves the Pareto optimal solution by the graph-based algorithm. This Pareto optimal solution simplifies the overall optimization problem and enables the central optimizer to solve it with smaller feasible regions