Greedy-based Method Research Articles

TRouter: Thermal-Driven PCB Routing via Nonlocal Crisscross Attention Networks

In this paper, we propose TRouter, a thermal-driven PCB routing framework via a machine learning model. The model is designed to capture the long-range spatial information from the PCB layout and predict thermal distribution. The information contains pads, vias, components and wire segments. A gradient in each grid cell obtained from the backpropagation is integrated into a full-board routing algorithm to guide thermal-aware wire detour and via punching. To achieve a significant speedup, we construct a conflict graph according to whether overlapping among convex hulls of nets. A greedy-based method is adopted to remove non-root nodes from all nodes. Then a task graph is constructed to improve the parallelism. We conduct experiments on open-source benchmarks to illustrate our TRouter can achieve significant speedup and lower-temperature designs, compared with a state-of-the-art PCB routing algorithm.

Double Deep Q-Network Based Dynamic Framing Offloading in Vehicular Edge Computing

With the rapid development of Artificial Intelligence (AI) and the Internet of Vehicles (IoV), there is an increasing demand for deploying various intelligent applications on vehicles. Vehicular Edge Computing (VEC) is receiving extensive attention from both the industry and academia due to its benefits from the edge computing paradigm, which pushes computing tasks from the core of the network to the edge of the network. However, in the VEC environment considering vehicles to Road Side Units (RSUs), due to the mobility of vehicles, it is still a challenge to make dynamic and efficient offloading decisions for compute-intensive tasks, especially in the congestion situation. In order to minimize the total delay and waiting time of tasks from moving vehicles, we establish a dynamic offloading model for multiple moving vehicles whose tasks can be divided into sequential subtasks, so that the offloading decisions are more refined. Moreover, the proposed model is frame-based to avoid unnecessary waiting time, which makes offloading decisions when the subtasks of each vehicle are generated rather than offloading subtasks after gathering subtasks of vehicles for a time slot. Aiming to find the optimal offloading decision for sequential subtasks, we propose a Dynamic Framing Offloading algorithm based on Double Deep Q-Network (DFO-DDQN). Extensive experimental results demonstrate the effectiveness and superiority of the proposed DFO-DDQN when compared with other DRL-based methods and greedy-based methods.

Improving supply quality in distribution power networks: A game-theoretic planning approach

Electric companies need to plan carefully the updates in their distribution power networks, assessing not only the expected impact on supply quality indicators, but also the cost of the different possibilities considered. To this end, we propose a planning method to improve the quality of supply in distribution power networks by upgrading devices so that they can be remotely operated. In the first place, our method identifies the lines with the lowest performance in an area of interest. Then, in each of these lines, the set of devices that can be upgraded is used to define the set of players of a cooperative game whose characteristic function provides a measure of the expected performance and economic impact of the updates in the infrastructure. Finally, the best locations to install remote control devices are determined by computing the Shapley value of this cooperative game, which, in addition, offers very valuable insight regarding each location considered. A case study based on a real substation in Spain is used to illustrate our approach and compare its performance with respect to the greedy-based method. • Design of a planning method to improve supply quality in distribution power networks. • Performance analysis of distribution lines based on a Spanish reliability index. • Use of game theory to find a trade-off between supply continuity and network updates. • Selection of the best locations to upgrade devices in the worst performance lines. • Assessment by a Spanish substation case study, outperforming greedy-based methods.

Intelligent-Reflecting-Surface-Aided Mobile Edge Computing With Binary Offloading: Energy Minimization for IoT Devices

Mobile edge computing (MEC) is envisioned as a promising technique to support computation-intensive and time-critical applications in future Internet of Things (IoT) era. However, the uplink transmission performance will be highly impacted by the hostile wireless channel, the low bandwidth, and the low transmission power of IoT devices. Recently, intelligent reflecting surface (IRS) has drawn much attention because of its capability to control the wireless environments so as to enhance the spectrum and energy efficiencies of wireless communications. In this article, we consider an IRS-aided multidevice MEC system where each IoT device follows the binary offloading policy, i.e., a task has to be computed as a whole either locally or remotely at the edge server. We aim to minimize the total energy consumption of devices by jointly optimizing the binary offloading modes, the CPU frequencies, the offloading powers, the offloading times, and the IRS phase shifts for all devices. Two algorithms, which are greedy based and penalty based, are proposed to solve the challenging nonconvex and discontinuous problem. It is found that the penalty-based method has only linear complexity with respect to the number of devices, but it performs close to the greedy-based method with cubic complexity with respect to the number of devices. Furthermore, binary offloading via IRS indeed saves more energy compared to the case without IRS.

A multi-objective worker selection scheme in crowdsourced platforms using NSGA-II

Crowdsourcing has led to a paradigm shift in how commercial houses execute projects by lowering the production cost. A crucial aspect of crowdsourcing is selecting the best set of workers to perform a task. The environment envisaged in this work is an independent pool of workers, each equipped with a pre-defined set of skills. We assume that these skills do not follow any priority order over each other. Given a task with a set of required skills, we aim to select a team of workers who can collectively fulfil the task’s requirements, while maximizing the collective expertise and minimizing the team cost. We propose a nondominated sorting genetic algorithm II (NSGA II) based algorithm to find the best set of workers that can perform the task. The basic operators of the evolutionary computation approach are tuned in accordance with our problem objectives. We perform a detailed analysis to show that the solution is well distributed along the Pareto optimal, converges exponentially apropos the number of generations and is cost-efficient. The proposed approach is compared with an optimal strategy, a greedy-based method, and another evolutionary-based algorithm to establish its effectiveness. Extensive simulation results using real data set and a synthetic data set were presented to validate our claim.

User Grouping and Reflective Beamforming for IRS-Aided URLLC

This letter studies an intelligent reflecting surface (IRS)-aided downlink ultra-reliable and low-latency communication (URLLC) system, in which an IRS is dedicatedly deployed to assist a base station (BS) to send individual short-packet messages to multiple users. To enhance the URLLC performance, the users are divided into different groups and the messages for users in each group are encoded into a single codeword. By considering the time division multiple access (TDMA) protocol among different groups, our objective is to minimize the total latency for all users subject to their individual reliability requirements, via jointly optimizing the user grouping and blocklength allocation at the BS together with the reflective beamforming at the IRS. We solve the latency minimization problem via the alternating optimization, in which the blocklengths and the reflective beamforming are optimized by using the techniques of successive convex approximation (SCA) and semi-definite relaxation (SDR), while the user grouping is updated by K-means and greedy-based methods. Numerical results show that the proposed designs can significantly reduce the communication latency, as compared to various benchmark schemes, which unveil the importance of user grouping and reflective beamforming optimization for exploiting the joint encoding gain and enhancing the worst-case user performance.

Identifying Influential Spreaders in Social Networks Through Discrete Moth-Flame Optimization

Influence maximization in a social network refers to the selection of node sets that support the fastest and broadest propagation of information under a chosen transmission model. The efficient identification of such influence-maximizing groups is an active area of research with diverse practical relevance. Greedy-based methods can provide solutions of reliable accuracy, but the computational cost of the required Monte Carlo simulations renders them infeasible for large networks. Meanwhile, although network structure-based centrality methods can be efficient, they typically achieve poor recognition accuracy. Here, we establish an effective influence assessment model based both on the total valuation and variance in valuation of neighbor nodes, motivated by the possibility of unreliable communication channels. We then develop a discrete moth-flame optimization method to search for influence-maximizing node sets, using a local crossover and mutation evolution scheme atop the canonical moth position updates. To accelerate convergence, a search area selection scheme derived from a degree-based heuristic is used. The experimental results on five real-world social networks, comparing our proposed method against several alternatives in the current literature, indicates our approach to be effective and robust in tackling the influence maximization problem.

A clique-based discrete bat algorithm for influence maximization in identifying top-k influential nodes of social networks

The problem of identifying the top-k influential node is still an open and deeply felt issue. The development of a stable and efficient algorithm to deal with such identification is still a challenging research hot spot. Although conventional centrality-based and greedy-based methods show high performance, they are not very efficient when dealing with large-scale social networks. Recently, algorithms based on swarm intelligence are applied to solve the problems mentioned above, and the existing researches show that such algorithms can obtain the optimal global solution. In particular, the discrete bat algorithm (DBA) has been proved to have excellent performance, but the evolution mechanism based on a random selection strategy leads to the optimal solution's instability. To solve this problem, in this paper, we propose a clique-DBA algorithm. The proposed algorithm is based on the clique partition of a network and enhances the initial DBA algorithm's stability. The experimental results show that the proposed clique-DBA algorithm converges to a determined local influence estimation (LIE) value in each run, eliminating the phenomenon of large fluctuation of LIE fitness value generated by the original DBA algorithm. Finally, the simulated results achieved under the independent cascade model show that the clique-DBA algorithm has a comparable performance of influence spreading compared with the algorithms proposed in the state of the art.

Receiver Design With Reduced DOF in Frequency Domain for Target Detection Under Gaussian Clutter

This paper addresses the problem of target detection against a background of Gaussian clutter by using frequency snapshots with reduced degrees of freedom (DOF). We derive the optimal detector and detection performance under the Neyman-Pearson criterion for general frequency snapshot selection with arbitrary DOF. When the clutter statistics are unknown, we use a uniformly random frequency snapshot selection method and show how the DOF employed affects the detection performance. When the clutter return follows a stationary Gaussian distribution with slowly varying power spectral density, the optimal selection is derived. When the clutter is composed of reflected versions of the transmitted waveforms, a greedy-based method for selecting the frequency snapshots is presented. Numerical experiments show that a receiver with reduced DOF can lead to detection performance which is very close to that of the receiver with full DOF.

A method for solving scriptio continua in Javanese manuscript transliteration

Many Javanese manuscripts in Indonesia are stored in museums and libraries. Most of these manuscripts were written using local scripts that are rarely used in everyday life, and hence a software application that can help and improve the reading of these manuscripts is valuable. An essential step in automatic manuscript image transliteration is post-processing, which involves editing and concatenating syllables into words. The main problem of post-processing is that there exists no symbol for space between words in a sentence, which is called the scriptio-continua problem. This paper proposes methods based on the backtracking algorithm to solve the scriptio continua in the post-processing step of Javanese manuscript image transliteration. The proposed methods use a depth-first search in seeking relevant candidate words to determine whether to merge a new syllable or not. The results of the proposed methods to concatenate 17,687 syllables from the Hamong Tani book using a dictionary containing 49,801 words are found to be satisfactory in terms of computation and accuracy. The accuracy of the implemented greedy and brute-force methods is both 81.64%. However, the greedy-based method is more efficient and has a better performance than the brute-force method.

Data Aggregation using Compressive Sensing for Energy Efficient Routing Strategy

Abstract Data gathering in an energy efficient way, is one of the crucial concerns in Wireless Sensor Networks (WSNs). With incorporation of Compressive Sensing (CS), as a data aggregation scheme the information contained in the signal, is safely maintained through its projections, which can be reconstructed later. Inclusion of CS, for an energy efficient routing technique further enhances the lifetime of the network. To improve network lifetime, CS has been employed at level 1 (at the leaf nodes). The dimensionality reduction at the transmitter is done, using a measurement matrix. At the receiver, data is recovered using greedy based methods. Greedy based method offers low complexity, and low implementation cost. The success rate of greedy method, depends on the sparsity of the data. Performance evaluation of greedy based method is analyzed, by considering varying sparsity and plotted against number of measurements, required to reconstruct the signal. At the same time, reconstruction error of every method has been evaluated by varying sparsity of the signal. The same is analyzed by considering real temperature, and humidity data sets.

Accelerated Best-First Search With Upper-Bound Computation for Submodular Function Maximization

Submodular maximization continues to be an attractive subject of study thanks to its applicability to many real-world problems. Although greedy-based methods are guaranteed to find (1-1/e)-approximate solutions for monotone submodular maximization, many applications require solutions with better approximation guarantees; moreover, it is desirable to be able to control the trade-off between the computation time and approximation guarantee. Given this background, the best-first search (BFS) has been recently studied as a promising approach. However, existing BFS-based methods for submodular maximization sometimes suffer excessive computation cost since their heuristic functions are not well designed. In this paper, we propose an accelerated BFS for monotone submodular maximization with a knapsack constraint. The acceleration is attained by introducing a new termination condition and developing a novel method for computing an upper-bound of the optimal value for submodular maximization, which enables us to use a better heuristic function. Experiments show that our accelerated BFS is far more efficient in terms of both time and space complexities than existing methods.

Efficient Computation of MSE Lower Bounds via Matching Pursuit

The classes of large-error bounds that are based on the covariance inequality, in both Bayesian and non-Bayesian approaches, are characterized as projection-based bounds. Tightening of bounds in these classes involves high computational complexity due to multidimensional optimization procedure. Consequently, projection-based large-error bounds have little popularity, while small-error bounds are frequently preferred, although they are not necessarily tight. In this letter, we first introduce a unified formulation for Bayesian and non-Bayesian projection-based lower bounds and set a general framework, which allows for their approximation via a greedy-based method. This framework is then used to propose the use of optimized orthogonal matching pursuit approach for computing projection-based large-error bounds. We analyze the complexity of the proposed algorithm and show that it is significantly lower than the complexity of the conventional approach. Finally, we apply the algorithm for the problem of multitone estimation and show that for fixed computational resources, the Weiss–Weinstein bound implemented with the proposed algorithm, provides a tighter bound compared to conventional approaches.

Energy management in Multi-Commodity Smart Energy Systems with a greedy approach

Along with the development of Smart Energy System (SES), the advancing popularity of hybrid energy appliances, such as micro-combined heat and power (CHP) and electric heaters, requires an overall energy management strategy to optimize the energy using while guaranteeing energy supply for both the electricity and heat demand. This paper based on the concept of Multi-Commodity Smart Energy System (MC-SES) proposes a self-sufficient hybrid energy model to minimize the energy exchange with the external electricity grid. Next to the optimization solution for this hybrid energy management problem, a greedy algorithm has been proposed to alleviate the high computational complexity of optimization searching. The optimization, direct allocation and greedy-based methods will all be analyzed, simulated and compared in this paper.

A new ant colony optimization-based algorithm for range query answering problem in relational schema-based P2P database systems

Nowadays, peer-to-peer database systems (P2PDBSs) aiming at data sharing in the Web have become very popular. Due to the absence of global knowledge about data placement in unstructured P2P networks, query processing and answering is a challenging problem in such systems. This process is provided by query routing and is an optimization problem whose goal is to find the maximum results with spending a predetermined cost. With the aim of improving the efficiency of range query answering algorithm in relational schema-based P2PDBSs, the present study, for the first time, adapts the ant colony metaphor with range query answering problem in relational schema-based P2P systems and proposes a new algorithm using the ant colony optimization approach in which the researchers use histogram data structure and apply both positive and negative feedbacks, dynamic learning rate, and local heuristic mechanisms and show that the proposed algorithm gives better results than the comparative greedy-based method. The experimental tests indicate that in the best case, the average number of traveled links for finding one answer (i.e., cost-to-answers ratio) is decreased almost by half in contrast to that of greedy-based algorithm. Furthermore, the achieved results indicate that the proposed algorithm is completely flexible with the users' requests, i.e., more query answers or less query response time, and the algorithm parameters can be properly set to meet the users' requests.

Topology Control for Time-Evolving and Predictable Delay-Tolerant Networks

In delay tolerant networks (DTNs), the lack of continuous connectivity, network partitioning, and long delays make design of network protocols very challenging. Previous DTN research mainly focuses on routing and information propagation. However, with a large number of wireless devices' participation, it becomes crucial regarding how to maintain efficient and dynamic topology of the DTN. In this paper, we study the topology control problem in a predictable DTN, where the time-evolving network topology is known a priori or can be predicted. We first model such time-evolving network as a directed space-time graph that includes both spacial and temporal information. The aim of topology control is to build a sparse structure from the original space-time graph such that 1) the network is still connected over time and supports DTN routing between any two nodes; 2) the total cost of the structure is minimized. We prove that this problem is NP-hard, and propose two greedy-based methods that can significantly reduce the total cost of topology while maintaining the connectivity over time. We also introduce another version of the topology control problem by requiring that the least cost path for any two nodes in this constructed structure is still cost-efficient compared with the one in the original graph. Two greedy-based methods are provided for such a problem. Simulations have been conducted on both random DTN networks and real-world DTN tracing data. Results demonstrate the efficiency of the proposed methods.