Alternating Optimization Approach Research Articles

Explainable Data Decompositions

Our goal is to discover the components of a dataset, characterize why we deem these components, explain how these components are different from each other, as well as identify what properties they share among each other. As is usual, we consider regions in the data to be components if they show significantly different distributions. What is not usual, however, is that we parameterize these distributions with patterns that are informative for one or more components. We do so because these patterns allow us to characterize what is going on in our data as well as explain our decomposition.We define the problem in terms of a regularized maximum likelihood, in which we use the Maximum Entropy principle to model each data component with a set of patterns. As the search space is large and unstructured, we propose the deterministic DISC algorithm to efficiently discover high-quality decompositions via an alternating optimization approach. Empirical evaluation on synthetic and real-world data shows that DISC efficiently discovers meaningful components and accurately characterises these in easily understandable terms.

Unsupervised deep hashing by joint optimization for pulmonary nodule image retrieval

Abstract In recent years, hash-based image retrieval has attracted great attention due to the rapid growth of medical images. In the paper, an end-to-end unsupervised deep hashing is proposed, where feature extraction and binary optimization are carried out by joint optimization. Our method consists of five components: a shared deep convolution neural network for learning image representations, a deconvolution module for reconstructing the original images, a classification module for leveraging semantic supervision by pseudo labels, a binary code learning module for encoding images features into binary codes, and a joint loss function for deep hash function learning. In addition, the real-valued features balanced in different dimensions by a rotation matrix are quantized directly into discrete binary codes in an alternating optimization approach to minimize the quantization loss. Experiments have been performed on the pulmonary nodule images dataset and the results demonstrate the proposed method can yield better retrieval performance by comparing with the state-of-the-art unsupervised hashing methods.

Joint Optimization of Relay Deployment, Channel Allocation, and Relay Assignment for UAVs-Aided D2D Networks

Unmanned aerial vehicles (UAVs) can be deployed in the air to provide high probabilities of line of sight (LoS) transmission, thus UAVs bring much gain for wireless communication systems. In this paper, we study a UAVs-aided self-organized device-to-device (D2D) network. Relay deployment, channel allocation and relay assignment are jointly optimized, aiming to maximize the capacity of the relay network. On account of the coupled relationship between the three optimization variables, an alternating optimization approach is proposed to solve this problem. The original problem is divided into two sub-problems. The first one is that of optimizing the channel allocation and relay assignment with fixed relay deployment. Considering without central controller, a reinforcement learning algorithm is proposed to solve this sub-problem. The second sub-problem is that of optimizing the relay deployment with fixed channel allocation and relay assignment. Assuming no knowledge of channel model and exact positions of the communication nodes, an online learning algorithm based on real-time capacity is proposed to solve this sub-problem. By solving the two sub-problems alternately and iteratively, the original problem is finally solved. Simulation results show that the UAVs-aided D2D network can achieve a high capacity via the joint optimization of relay deployment, channel allocation, and relay assignment.

Joint Trajectory and Resource Allocation Design for Energy-Efficient Secure UAV Communication Systems

In this paper, we study the trajectory and resource allocation design for downlink energy-efficient secure unmanned aerial vehicle (UAV) communication systems, where an information UAV assisted by a multi-antenna jammer UAV serves multiple ground users in the existence of multiple ground eavesdroppers. The resource allocation strategy and the trajectory of the information UAV, and the jamming policy of the jammer UAV are jointly optimized for maximizing the system energy efficiency. The joint design is formulated as a non-convex optimization problem taking into account the quality of service (QoS) requirement, the security constraint, and the imperfect channel state information (CSI) of the eavesdroppers. The formulated problem is generally intractable. As a compromise approach, the problem is divided into two subproblems which facilitates the design of a low-complexity suboptimal algorithm based on alternating optimization approach. Simulation results illustrate that the proposed algorithm converges within a small number of iterations and demonstrate some interesting insights: (1) the introduction of a jammer UAV facilitates a highly flexible trajectory design of the information UAV which is critical to improving the system energy efficiency; (2) by exploiting the spatial degrees of freedom brought by the multi-antenna jammer UAV, our proposed design can focus the artificial noise on eavesdroppers offering a strong security mean to the system.

Spoken Language Identification Based on Particle Swarm Optimisation–Extreme Learning Machine Approach

The determination and classification of natural language based on specified content and data set involves a process known as spoken language identification (LID). To initiate the process, useful features of the given data need to be extracted first in a mature process where the standard LID features have been previously developed by employing the use of MFCC, SDC, GMM and the i-vector-based framework. Nevertheless, optimisation of the learning process is still required to enable a comprehensive capturing of the extracted features’ embedded knowledge. The training of a single hidden layer neural network can be done using the extreme learning machine (ELM), which is an effective learning model for conducting classification and regression analysis. Nevertheless, the learning process of this model is not entirely effective (i.e. optimised) due to the random selection of weights within the input hidden layer. This study employs ELM as the LID learning model centred upon the extraction of the standard features. The enhanced self-adjusting extreme learning machine (ESA–ELM) is one of the ELM’s optimisation techniques which has been chosen as the benchmark and is enhanced by adopting a new alternative optimisation approach (PSO) instead of (EATLBO) in terms of achieving high performance. The improved ESA–ELM is named particle swarm optimisation–extreme learning machine (PSO–ELM). The generated results are based on LID with the same benchmarked data set derived from eight languages, which indicated the superior performance of the particle swarm optimisation–extreme learning machine LID (PSO–ELM LID) with an accuracy of 98.75% in comparison with the ESA–ELM LID which only achieved 96.25%.

CD19-Loss with Preservation of Other B Cell Lineage Features in Patients with Large B Cell Lymphoma Who Relapsed Post-Axi-Cel

CD19-Loss with Preservation of Other B Cell Lineage Features in Patients with Large B Cell Lymphoma Who Relapsed Post-Axi-Cel

Joint design of beamforming and time switching/power splitting for wireless-powered multi-antenna dual-relay network

In this paper, we consider a wireless-powered dual-relay network consisting of one multi-antenna source, two single-antenna energy-constrained relays and one single-antenna destination without direct source to destination link. In order to establish the communication flow, the energy-constrained relays harvest energy from the radio frequency transmitted by the source firstly, then exploit the harvested energy to forward the source information to the destination based on distributed space time coding (DSTC). Under this network architecture, three decode-and-forward (DF) technique-based relaying protocols, i.e., time switching-based relaying (TSR) protocol, power splitting-based relaying (PSR) protocol, and hybrid relaying (HR) protocol, are considered to drive the energy transfer and information transmission. To maximize the network throughput, the joint design for the optimal energy and information beamforming vectors employed at the source, the optimal time switching, and power splitting ratios under these three protocols are investigated and solved efficiently by employing simple sequential optimization approach or alternating optimization approach. Simulations are conducted to show the superior performance achieved by our proposed scheme. Moreover, we find that the TSR protocol outperforms the PSR protocol in the low signal-to-noise ratio (SNR) region, while the latter outperforms the former in the high SNR region. And the HR protocol achieves the best performance in any SNR region. At the same time, the effect of the relays’ locations on the throughput performance of these three protocols is also investigated.

Hybrid Beamforming With Finite-Resolution Phase Shifters for Multiuser Millimeter-Wave Downlink

This letter considers hybrid analog-digital beamforming with discrete-phase shifters for multiuser multiple-input and single-output (MISO) downlink transmission in millimeter-wave (mm-wave) channels. A max-min signal-to-interference-plus-noise ratio (SINR) fairness problem is formulated. Owing to the discrete-phase constraints, this problem is a mixed integer nonlinear program (MINP) and is generally NP-hard. To handle it, an alternating optimization approach is employed. Specifically, for the digital beamformer, zero-forcing (ZF) beamforming is utilized. For the discrete-phase analog beamformer, we first derive an equivalent continuous reformulation and then solve it by leveraging an inexact majorization-minimization (MM) algorithm, which is built upon the first-order method and MM update. Simulation results show that the proposed design can almost achieve the optimal performance of the exhaustive search method, and can approach the optimal full-digital performance with increasing resolution of the discrete-phase shifters.

Manta ray foraging optimization: An effective bio-inspired optimizer for engineering applications

A new bio-inspired optimization technique, named Manta Ray Foraging Optimization (MRFO) algorithm, is proposed and presented, aiming to providing a novel algorithm that provides an alternate optimization approach for addressing real-world engineering issues. The inspiration of this algorithm is based on intelligent behaviors of manta rays. This work mimics three unique foraging strategies of manta rays, including chain foraging, cyclone foraging, and somersault foraging, to develop an efficient optimization paradigm for solving different optimization problems. The performance of MRFO is evaluated, through comparisons with other state-of-the-art optimizers, on benchmark optimization functions and eight real-world engineering design cases. The comparison results on the benchmark functions suggest that MRFO is far superior to its competitors. In addition, the real-world engineering applications show the merits of this algorithm in tackling challenging problems in terms of computational cost and solution precision. The MATLAB codes of the MRFO algorithm are available at https://www.mathworks.com/matlabcentral/fileexchange/73130-manta-ray-foraging-optimization-mrfo.

Faster algorithms for min-max-min robustness for combinatorial problems with budgeted uncertainty

We consider robust combinatorial optimization problems where the decision maker can react to a scenario by choosing from a finite set of k solutions. This approach is appropriate for decision problems under uncertainty where the implementation of decisions requires preparing the ground. We focus on the case that the set of possible scenarios is described through a budgeted uncertainty set and provide three algorithms for the problem. The first algorithm solves heuristically the dualized problem, a non-convex mixed-integer non-linear program (MINLP), via an alternating optimization approach. The second algorithm solves the MINLP exactly for k=2 through a dedicated spatial branch-and-bound algorithm. The third approach enumerates k-tuples, relying on strong bounds to avoid a complete enumeration. We test our methods on shortest path instances that were used in the previous literature and on randomly generated knapsack instances, and find that our methods considerably outperform previous approaches. Many instances that were previously not solved within hours can now be solved within few minutes, often even faster.

Single-arc VMAT optimization for dual-layer MLC

Dual-layer multi-leaf collimator (DLMLC) has recently attracted renewed interest due to its good balance among resolution, low leakage, and high fabricability. However, existing progressive sampling based volumetric modulated arc therapy (VMAT) algorithm is ineffective for DLMLC, requiring more arcs to achieve dosimetry comparable to VMAT plans with higher resolution single-layer MLC (SLMLC). In this study, we develop a novel single-arc VMAT optimization framework to take advantage of the unique DLMLC characteristics fully. Direct aperture optimization (DAO) for single-arc DLMLC VMAT was formulated as a least square dose fidelity objective, along with an anisotropic total variation term to regulate the fluence smoothness and a single segment term for forming simple apertures. The DAO was solved through alternating optimization approach. The DLMLC deliverability constraint and the MLC leaf speed constraint were formulated as the optimization constraints and solved using a graph optimization algorithm. Feasibility of the proposed framework was tested on a brain, a lung, and a prostate cancer patient. The framework was further adapted for a simultaneous integrated boost (SIB) case. The single-arc DLMLC-10 mm (leaf width) plan was compared against single-arc SLMLC VMAT plans including SLMLC-5mm, SLMLC-10mm, and SLMLC with 10 mm leaf width and 5 mm leaf step size (SLMLC-10mm-5mm). Compared with the SLMLC-10mm plan and the SLMLC-10mm-5mm plan, with the same target coverage, the DLMLC-10 mm plan reduced R50 by 30.7% and 10.0%, the average max OAR dose by 5.79% and 3.7% of the prescription dose, and the average mean OAR dose by 4.18% and 2.1% of the prescription dose, respectively. The plan quality is comparable to that of the SLMLC-5mm plan. The novel single-arc VMAT optimization framework for DLMLC utilizes two MLC layers to improve the effective modulation resolution and afford more sophisticated modulation. Consequently, DLMLC VMAT achieves superior dosimetry to SLMLC VMAT with the same leaf width.

Improved multiobjective salp swarm optimization for virtual machine placement in cloud computing

In data center companies, cloud computing can host multiple types of heterogeneous virtual machines (VMs) and provide many features, including flexibility, security, support, and even better maintenance than traditional centers. However, some issues need to be considered, such as the optimization of energy usage, utilization of resources, reduction of time consumption, and optimization of virtual machine placement. Therefore, this paper proposes an alternative multiobjective optimization (MOP) approach that combines the salp swarm and sine-cosine algorithms (MOSSASCA) to determine a suitable solution for virtual machine placement (VMP). The objectives of the proposed MOSSASCA are to maximize mean time before a host shutdown (MTBHS), to reduce power consumption, and to minimize service level agreement violations (SLAVs). The proposed method improves the salp swarm and the sine-cosine algorithms using an MOP technique. The SCA works by using a local search approach to improve the performance of traditional SSA by avoiding trapping in a local optimal solution and by increasing convergence speed. To evaluate the quality of MOSSASCA, we perform a series of experiments using different numbers of VMs and physical machines. The results of MOSSASCA are compared with well-known methods, including the nondominated sorting genetic algorithm (NSGA-II), multiobjective particle swarm optimization (MOPSO), a multiobjective evolutionary algorithm with decomposition (MOEAD), and a multiobjective sine-cosine algorithm (MOSCA). The results reveal that MOSSASCA outperforms the compared methods in terms of solving MOP problems and achieving the three objectives. Compared with the other methods, MOSSASCA exhibits a better ability to reduce power consumption and SLAVs while increasing MTBHS. The main differences in terms of power consumption between the MOSCA, MOPSO, MOEAD, and NSGA-II and the MOSSASCA are 0.53, 1.31, 1.36, and 1.44, respectively. Additionally, the MOSSASCA has higher MTBHS value than MOSCA, MOPSO, MOEAD, and NSGA-II by 362.49, 274.70, 585.73 and 672.94, respectively, and the proposed method has lower SLAV values than MOPSO, MOEAD, and NSGA-II by 0.41, 0.28, and 1.27, respectively.

Enhancing Performance of Random Caching in Large-Scale Wireless Networks With Multiple Receive Antennas

To improve signal-to-interference ratio (SIR) and make better use of file diversity provided by random caching, we consider two types of linear receivers, i.e., maximal ratio combining (MRC) receiver and partial zero forcing (PZF) receiver, at users in a large-scale cache-enabled single-input multi-output network. First, for each receiver, by utilizing tools from stochastic geometry, we derive a tractable expression and a tight upper bound for the successful transmission probability (STP). In the case of the MRC receiver, we also derive a closed-form expression for the asymptotic outage probability in the low SIR threshold regime. Then, for each receiver, we maximize the STP. In the case of the MRC receiver, we consider the maximization of the tight upper bound on the STP by optimizing the caching probabilities, which is a non-convex problem. We obtain a stationary point, by solving an equivalent difference of convex programming problem using concave-convex procedure. We also obtain a closed-form asymptotically optimal solution in the low SIR threshold regime. In the case of the PZF receiver, we consider the maximization of the tight upper bound on the STP by optimizing the caching probabilities and the degrees of freedom allocation (for boosting signal power and canceling interference), which is a mixed discrete-continuous problem. Based on structural properties, we obtain a low-complexity near optimal solution by using an alternating optimization approach. The analysis and optimization results reveal the impact of antenna resource at users on random caching.

Recover feasible solutions for SOCP relaxation of optimal power flow problems in mesh networks

The AC optimal power flow (OPF) problem is essential for the schedule and operation of power systems. Convex relaxation methods have been studied and used extensively to obtain an optimal solution to the OPF problem. When the exactness of convex relaxations is not guaranteed, it is important to recover a feasible solution for the convex relaxation methods. This paper presents an alternative convex optimisation (ACP) approach that can efficiently recover a feasible solution from the result of second-order cone programming (SOCP) relaxed OPF in mesh networks. The OPF problem is first formulated as a difference-of-convex programming (DCP) problem, then efficiently solved by a penalty convex concave procedure (CCP). By using the solution of a tightened SOCP OPF as an initial point, the proposed algorithm is able to find a global or near-global optimal solution to the AC OPF problem. Numerical tests show that the proposed method outperforms those semi-definite programming (SDP) and quadratically constrained quadratic programming (QCQP)-based algorithms.

Active Transfer Learning

A major assumption in data mining and machine learning is that the training set and test set come from the same domain. They share the same feature space and have the same distribution. However, in many real-world applications, the training set and test set usually come from different domains. Thus, there might be negative similarities between different domains so that the negative transfer problem caused by negative similarity may happen. In this paper, we propose a novel method named active transfer learning (ATL) to solve the above problem. Specifically, the orthogonal projection matrix and the weight coefficient vector are introduced to extend maximum mean discrepancy (MMD) so that it can minimize MMD and simultaneously eliminate the negative transfer. To find the informative and discriminative subsets from the source domain, we then propose an information diversity term by using the local geometric structure information of the source samples. Besides, by using the label information of source samples, our method can guarantee the selected subsets as discriminative as possible. Finally, to efficiently implement the proposed method, an alternating optimization approach, which is based on the alternating direction method of multipliers (ADMM), is designed to solve the optimization problem. To demonstrate the effectiveness of the proposed ATL model, experiments are conducted on five real-world data sets. The experimental results show the superiority of our method over the state-of-the-art methods.

A Surrogate Optimization-Based Mechanism for Resource Allocation and Routing in Networks With Strategic Agents

We consider a mechanism design problem for the joint flow control and multipath routing in informationally decentralized networks with strategic agents. Based on a surrogate optimization approach, we propose an incentive mechanism that strongly implements the social-welfare maximizing outcome in Nash equilibria. This mechanism possesses several other desirable properties, including individual rationality and budget balance at equilibrium. More importantly, in contrast to the existing literature on the network resource allocation mechanisms, the proposed mechanism is dynamically stable, meaning that the Nash equilibrium (NE) of the game induced by the mechanism can be learned by the agents in a decentralized manner. To establish dynamic stability, we propose a decentralized iterative process that always converges to a NE of the game induced by the mechanism, provided that all strategic agents follow the process. To the best of our knowledge, this is the first incentive mechanism that simultaneously possesses all the above-mentioned properties.

Cooperative SWIPT Transmission in Overlay Multi-Antennas Hierarchical Cognitive Radio Networks

In this paper, we propose a new multi-antennas hierarchical cognitive radio, in which a primary system (PS) not only shares the spectrum resources with a secondary system (SS), but also harvests energy from the SS, achieving a new simultaneous wireless information and power transfer paradigm. Specifically, the secondary transmitter (ST) can help relay PS’s signals via an amplify-and-forward protocol, and a power splitting energy harvesting device is additionally adopted at the primary receiver (PR). A cooperative strategy is proposed for maximizing the achievable rate of the SS while ensuring PS’s achievable rate and PR’s harvested energy, by jointly optimizing a precoder at the primary transmitter, two precoders at the ST (for the PS and SS, respectively), an equalizer at the secondary receiver, and a power splitting factor for the PR. Since the optimization problem is complicated and non-convex, we propose an alternating optimization approach to decompose the original problem into three subproblems, which can finally be formulated as semi-definite programming, difference of convex, and generalized eigenvalue problems. The closed-form solutions are, then, tractably derived. The numerical results are carried out to demonstrate the effectiveness of the proposed design.

Secrecy Rate Maximization for Untrusted Relay Networks With Nonorthogonal Cooperative Transmission Protocols

In this paper, a relay based network is considered where the source transmits confidential message to the destination via a nonregenerative relay. Due to the fact that the relay is unidentified, and thus may wiretap the relayed message from source to destination. In this regard, two new half-duplex nonorthogonal amplify-and-forward (NAF) secure communication protocols are proposed, with different assumptions on the transmission ability of the destination, i.e., NAF-I for destination without transmission ability and NAF-II for destination with transmission ability. We separately derive the sufficient and necessary conditions to yield positive secrecy rate for these protocols under fixed source/destination transmission power, and discuss their joint power allocation that maximizes the system secrecy rate. Despite the nonconvexity of the considered power allocation problems, alternative optimization approach is proved useful in dealing with these problems, and the corresponding algorithms are proposed. We further prove the convergence of the proposed algorithms and analyze their complexity. The discussion is also extended to the orthogonal amplify-and-forward (OAF) based protocols, i.e., OAF-I or OAF-II, for comparison purpose. Our numerical results demonstrate the gain of the secrecy rate performance of the proposed NAF protocols over existing transmission protocols, and also verify the effectiveness of the designed alternative optimization algorithms.

Symbol-Level Precoding Design Based on Distance Preserving Constructive Interference Regions

In this paper, we investigate the symbol-level precoding (SLP) design problem in the downlink of a multiuser multiple-input single-output (MISO) channel. We consider generic constellations with any arbitrary shape and size, and confine ourselves to one of the main categories of constructive interference regions (CIR), namely, distance preserving CIR (DPCIR). We provide a comprehensive study of DPCIRs and derive some properties for these regions. Using these properties, we first show that any signal in a given DPCIR has a norm greater than or equal to the norm of the corresponding constellation point if and only if the convex hull of the constellation contains the origin. It is followed by proving that the power of the noiseless received signal lying on a DPCIR is a monotonic strictly increasing function of two parameters relating to the infinite Voronoi edges. Using the convex description of DPCIRs and their properties, we formulate two design problems, namely, the SLP power minimization with signal-to-interference-plus-noise ratio (SINR) constraints, and the SLP SINR balancing problem under max-min fairness criterion. The SLP power minimization based on DPCIRs can straightforwardly be written as a quadratic program (QP). We provide a simplified reformulation of this problem which is less computationally complex. The SLP max-min SINR, however, is non-convex in its original form, and hence difficult to tackle. We propose several alternative optimization approaches, including semidefinite program (SDP) formulation and block coordinate descent (BCD) optimization. We discuss and evaluate the loss due to the proposed alternative methods through extensive simulation results.

Transmit Waveform/Receive Filter Design for MIMO Radar With Multiple Waveform Constraints

In this paper, we consider the joint design of both transmit waveforms and receive filters for a colocated multiple-input-multiple-output (MIMO) radar with the existence of signal-dependent interference and white noise. The design problem is formulated into a maximization of the signal-to-interference-plus-noise ratio (SINR), including various constraints on the transmit waveforms. Compared with the traditional alternating semidefinite relaxation approach, a general and flexible algorithm is proposed based on the majorization-minimization method with guaranteed monotonicity, lower computational complexity per iteration and/or convergence to a B-stationary point. Many waveform constraints can be flexibly incorporated into the algorithm with only a few modifications. Furthermore, the connection between the proposed algorithm and the alternating optimization approach is revealed. Finally, the proposed algorithm is evaluated via numerical experiments in terms of SINR performance, ambiguity function, computational time, and properties of the designed waveforms. The experiment results show that the proposed algorithms are faster in terms of running time and meanwhile achieve as good SINR performance as the the existing methods.