Optimizing Gateway Placement in Wireless Mesh Networks Based on ACO Algorithm

Abstract

In this paper, we study the challenging problem ofoptimizing gateway placement for throughput in Wireless MeshNetworks and propose a novel algorithm based on Ant ColonyOptimization (ACO) for it. The ACO algorithm is originatedfrom ant behavior in the food searching based on pheromone.We generate the locations of gateway randomly andindependently then calculate the probability and pheromonevalues of ants will choose to go from current gateway i to nextclient j. After each iteration, the pheromone values are updatedby all the number of ants that have reached to the destinationsuccessfully and found a optimal solution. Numerical resultsshow that the proposed algorithm has achieved much betterthan previous studies

Optimal Planning of Emergency Communication Network Using Deep Reinforcement Learning

“…Traditional planning algorithms require excessive amounts of prior knowledge and take too long to produce results. Accordingly, this paper draws on the intelligent decision ideas of AlphaZero [5], comprehensively considers the two factors of network coverage and connectivity, and introduces deep reinforcement learning algorithms [6], [7] to solve topology planning in emergency communication networks. First, according to the characteristics of the emergency communication network, learn from chess game ideas, and establish an abstract model of network planning, which realize the abstract mapping of the emergency communication network/node to the chessboard/chess.…”