Bianca de Almeida Dantas

Prof. Dr. Bianca de Almeida Dantas

orcid

Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 October 2018 in Journal of Parallel and Distributed Computing
Verified
Journal Article
Closed access
BDBianca de Almeida Dantas
Bianca de Almeida Dantas
ECEdson Norberto Cáceres
Edson Norberto Cáceres

1 October 2016
Verified
Conference Paper
Closed access
BDBianca de Almeida Dantas
Bianca de Almeida Dantas
ECEdson Norberto Caceres
Edson Norberto Caceres
In the last decades, with the advances in multicore/manycore architectures, it became interesting to design algorithms which can take advantage of such architectures aiming the achievement of more efficient algorithms to solve difficult problems. A large number of real-world problems solved with the help of computer programs demand faster or better quality solutions. Some of these problems can be modeled as classical theoretical problems, such as the 0-1 multidimensional knapsack problem (0-1 MKP), known to belong to the NP-hard class of problems, for which we can not obtain an exact solution efficiently. This motivates the search for alternative strategies which can achieve good quality approximate solutions, like metaheuristics, and also different ways to enable their execution in reduced times, such as parallel algorithms which explore multicore/manycore architectures. In this work we describe a parallelization of a simulated annealing approachusing GPGPU to solve 0-1 MKP and compare our results to previous works in order to prove the viability of its use.

1 January 2015 in Procedia Computer Science
Verified
Journal Article
Open access
BDBianca de Almeida Dantas
Bianca de Almeida Dantas
ECEdson Norberto Cáceres
Edson Norberto Cáceres
The knapsack problem is a widely known problem in combinatorial optimization and has been object of many researches in the last decades. The problem has a great number of variants and obtaining an exact solution to any of these is not easily accomplished, which motivates the search for alternative techniques to solve the problem. Among these alternatives, metaheuristics seem to be suitable on the search for approximate solutions for the problem. In this work we propose a sequential and a parallel implementation for the multidimensional knapsack problem using GRASP metaheuristic. The obtained results show that GRASP can lead to good quality results, even optimal in some instances, and that CUDA may be used to expand the neighborhood search and as a result may lead to improved quality results

1 September 2014
Verified
Conference Paper
Closed access
BDBianca de Almeida Dantas
Bianca de Almeida Dantas
ECEdson Norberto Caceres
Edson Norberto Caceres
The knapsack problem is a widely known problem in combinatorial optimization and has been object of many researches in the last decades. The problem has a great number of variants and obtaining an exact solution to any of these is not easily accomplished, which motivates the search for alternative techniques to solve the problem. Among these alternatives, augmented neural networks seem to be suitable on the search for approximate solutions for the problem. In this work we propose a parallel implementation for the multidimensional knapsack problem using augmented neural networks. The obtained results show that augmented neural networks allow efficient parallelization using CUDA: even smaller numbers of epoques resulted on equal or even better solutions than the sequential implementation. Differently, MPI implementation did not achieve satisfactory results.

19 May 2025 in Revista Principia
Verified
Journal Article
Open access
PDF file
WCWilton Gustavo Gomes da Costa
Wilton Gustavo Gomes da Costa
RSRicardo Santos
Ricardo Santos
WSWilly Alves de Oliveira Soler
Willy Alves de Oliveira Soler
BDBianca de Almeida Dantas
Bianca de Almeida Dantas
The Vehicle Routing Problem (VRP) is a classical optimization problem focused on determining routes for a fleet of vehicles to fulfill the demands of a set of customers. Due to its wide applicability, particularly in the logistics sector, numerous variants of the VRP have been developed. One such variant is the Dynamic Vehicle Routing Problem (DVRP), characterized by the emergence of new customer demands after the initial routing has been established. The combinatorial complexity of the DVRP poses significant challenges in finding high-quality, feasible solutions within practical runtime limits for real-world, large-scale problems, thereby driving the development of heuristic approaches. This paper introduces a novel heuristic algorithm, Dynamic Search per Neighbors Routes (DSNR), which employs neighborhood search techniques based on the 2-Opt* operator to incorporate new delivery requests into existing routes. The DSNR algorithm was tested using instances from the Loggibud benchmark, representing real-world data from three distinct Brazilian cities. In terms of distance and the number of vehicles required, the proposed approach demonstrated superior performance compared to two existing methods from the literature: the QRP Sweep (QRPS) and K-Means Greedy (KG). The DSNR achieved improvements of 15% to 17% in terms of distance and final delivery costs.

14 August 2026 in Scientific Data
Verified
Journal Article
Open access
PDF file
MTMurilo Taparo
Murilo Taparo
BDBianca de Almeida Dantas
Bianca de Almeida Dantas
RSRicardo Santos
Ricardo Santos
This paper presents the Network and Application Dataset for Internet of Things (NAD-IoT [1]), a dataset built to support research on job allocation in large-scale IoT networks. The dataset provides instances from 10 (101) up to 1,000,000,000 (109) network nodes, each associated with multiple jobs that represent application demands. Nodes and jobs of the dataset have essential attributes such as processing capacity, bandwidth, and latency, allowing researchers to explore performance under a wide range of operating conditions. NAD-IoT is organized into scenarios defined by the ratio between jobs and nodes: Lightweight scenarios, where jobs represent up to 20% of the nodes, and Heavyweight scenarios, where jobs correspond to 70% of the nodes. It enables systematic assessments of heuristics, metaheuristics, and optimization strategies, supporting analyses of scalability, efficiency, and runtime. It provides a reliable foundation for benchmarking algorithmic behavior, examining job-distribution patterns, and evaluating resource-management techniques. Beyond optimization studies, NAD-IoT can be used for experiments, simulation design, model training, and validation in investigations involving large-scale, heterogeneous, and resource-constrained IoT environments.