Alternating Optimization Approach Research Articles

An integrated dimensionality reduction and surrogate optimization approach for plant‐wide chemical process operation

AbstractWith liquefied natural gas becoming increasingly prevalent as a flexible source of energy, the design and optimization of industrial refrigeration cycles becomes even more important. In this article, we propose an integrated surrogate modeling and optimization framework to model and optimize the complex CryoMan Cascade refrigeration cycle. Dimensionality reduction techniques are used to reduce the large number of process decision variables which are subsequently supplied to an array of Gaussian processes, modeling both the process objective as well as feasibility constraints. Through iterative resampling of the rigorous model, this data‐driven surrogate is continually refined and subsequently optimized. This approach was not only able to improve on the results of directly optimizing the process flow sheet but also located the set of optimal operating conditions in only 2 h as opposed to the original 3 weeks, facilitating its use in the operational optimization and enhanced process design of large‐scale industrial chemical systems.

High-dimensional black-box optimization under uncertainty

Optimizing expensive black-box systems with limited data is an extremely challenging problem. As a resolution, we present a new surrogate optimization approach by addressing two gaps in prior research—unimportant input variables and inefficient treatment of uncertainty associated with the black-box output. We first design a new flexible non-interpolating parsimonious surrogate model using a partitioning-based multivariate adaptive regression splines approach, Tree Knot MARS (TK-MARS). The proposed model is specifically designed for optimization by capturing the structure of the function, bending at near-optimal locations, and is capable of screening unimportant input variables. Furthermore, we develop a novel replication approach called Smart-Replication, to overcome the uncertainty associated with the black-box output. The Smart-Replication approach identifies promising input points to replicate and avoids unnecessary evaluations of other data points. Smart-Replication is agnostic to the choice of a surrogate and can adapt itself to an unknown noise level. Finally to demonstrate the effectiveness of our proposed approaches we consider different complex global optimization test functions from the surrogate optimization literature. The results indicate that TK-MARS outperforms original MARS within a surrogate optimization algorithm and successfully detects important variables. The results also show that although non-interpolating surrogates can mitigate uncertainty, replication is still beneficial for optimizing highly complex black-box functions. The robustness and the quality of the final optimum solution found through Smart-Replication are competitive with that using no replications in environments with low levels of noise and using a fixed number of replications in highly noisy environments.

An alternative approach for bioanalytical assay optimization for wastewater-based epidemiology of SARS-CoV-2

Wastewater-based epidemiology of SARS-CoV-2 could play a role in monitoring the spread of the virus in the population and controlling possible outbreaks. However, sensitive sample preparation and detection methods are necessary to detect trace levels of SARS-CoV-2 RNA in influent wastewater (IWW).Unlike predecessors, method optimization of a SARS-CoV-2 RNA concentration and detection procedure was performed with IWW samples with high viral SARS-CoV-2 RNA loads. This is of importance since the SARS-CoV-2 genome in IWW might have already been subject to in-sewer degradation into smaller genome fragments or might be present in a different form (e.g. cell debris, …). Centricon Plus-70 (100 kDa) centrifugal filter devices resulted in the lowest and most reproducible Ct-values for SARS-CoV-2 RNA. Lowering the molecular weight cut-off did not improve our limit of detection and quantification (approximately 100 copies/μL for all genes). Quantitative polymerase chain reaction (qPCR) was employed for the amplification of the N1, N2, N3 and E-gene fragments.This is one of the first studies to apply digital polymerase chain reaction (dPCR) for the detection of SARS-CoV-2 RNA in IWW. dPCR showed high variability at low concentration levels (100 copies/μL), indicating that variability in bioanalytical methods for wastewater-based epidemiology of SARS-CoV-2 might be substantial. dPCR results in IWW were in line with the results found with qPCR.On average, the N2-gene fragment showed high in-sample stability in IWW for 10 days of storage at 4 °C. Between-sample variability was substantial due to the low native concentrations in IWW. Additionally, the E-gene fragment proved to be less stable compared to the N2-gene fragment and showed higher variability. Freezing the IWW samples resulted in a 10-fold decay of loads of the N2- and E-gene fragment in IWW.

AN ALTERNATIVE FRAMEWORK FOR SOCIALLY RESPONSIBLE PORTFOLIOS OPTIMIZATION APPLIED TO THE MOROCCAN STOCK EXCHANGE

The purpose of this article is to propose an alternative approach for portfolio optimization combining financial and ethical constraints in one hand, and objective and subjective investor’s preferences in the other hand. This approach intends to support investors in selecting and optimizing financial and social portfolio’s performances. More precisely, we introduce analytic hierarchy process (AHP) to measure the ethical performance (EP) score of each asset considering ethical criteria. For its part, fuzzy multiple criteria decision making (FMCDM) is used to determine the overall financial quality score of assets with respect to key financial criteria i.e., short term return, long term return, and risk. Besides, interactive fuzzy programming approach is applied to support the investor’s decision, considering its subjective preferences. The robustness of our approach is tested through an empirical study involving the case of the Casablanca Stock Exchange (CSE). The results give evidence that the Socially Responsible (SR) Portfolio have performed similar to the conventional one, as no significant differences were found in term of return. However, the SR portfolio allows the investor to achieve his ethical goal, against a slight financial sacrifice.

Secure IoT Communications Using HARQ-Based Beamforming for MISOSE Channels

In this article, we consider physical-layer security (PLS) of the Internet-of-Things (IoT) communications based on hybrid automatic repeat request (HARQ) protocols in multiple-input single-output single-eavesdropper (MISOSE) channels. Owing to the reciprocity of wireless channels, a negative acknowledge frame can be used to obtain the channel state information (CSI) from a source with multiple antennas to an IoT destination device, and this CSI of a legitimate link is used for beamforming for the next transmission round (TR). Therefore, a higher secrecy rate can be expected from retransmissions with beamforming based on the CSI as compared to the first TR without CSI. To evaluate the secrecy throughput under time-invariant and time-varying channels, connection and secrecy outage probabilities (OPs) are derived. By exploiting the characteristics of the OPs, we suggest an approximated secrecy throughput. We then consider the optimization of the approximated secrecy throughput considering the rate parameters for legitimate and eavesdropping links. With an alternating optimization approach, an algorithm to find convergent system parameters, i.e., the rates for legitimate and eavesdropping links, is proposed. Moreover, an asymptotic analysis with the large number of transmit antennas indicates an interesting behavior of the secrecy rate with regard to the number of TRs. The analytical results are verified by rigorous numerical simulations and they can serve as a useful guideline to those involved in the design of secure IoT communication systems with HARQ in MISOSE channels.

Reducing self-shading effects in Botryococcus braunii cultures: effect of Mg2+ deficiency on optical and biochemical properties, photosynthesis and lipidomic profile

Microalgae biomass exploitation as a carbon–neutral energy source is currently limited by several factors, productivity being one of the most relevant. Due to the high absorption properties of light-harvesting antenna, photosynthetic cells tend to capture an excessive amount of energy that cannot be entirely channeled through the electron transfer chain that ends up dissipated as heat and fluorescence, reducing the overall light use efficiency. Aiming to minimize this hurdle, in this work we studied the effect of decreasing concentrations of Magnesium (Mg2+) on the chlorophyll a content, photosynthetic performance, biomass and lipid production of autotrophic cultures of Botryococcus braunii LB 572. We also performed, for the first time, a comparative lipidomic analysis to identify the influence of limited Mg2+ supply on the lipid profile of this algae. The results indicated that a level of 0.0037 g L−1 MgSO4 caused a significant decline on chlorophyll a content with a concomitant 2.3-fold reduction in the biomass absorption coefficient. In addition, the Mg2+ limitation caused a decrease in the total carbohydrate content and triggered lipid accumulation, achieving levels of up to 53% DCW, whereas the biomass productivity remained similar for all tested conditions. The lipidome analysis revealed that the lowest Mg2+ concentrations also caused a differential lipid profile distribution, with an enrichment of neutral lipids and an increase of structural lipids. In that sense, we showed that Mg2+ limitation represents an alternative optimization approach that not only enhances accumulation of neutral lipids in B. braunii cells but also may potentially lead to a better areal biomass productivity due to the reduction in the cellular light absorption properties of the cells.

Constant Envelope Precoding With Extended Degrees of Freedom Through Per-User Symbol Scaling

In this letter, we propose a novel constant envelope precoding (CEP) design for multi-user multi-input single-output (MU-MISO) downlink, where we exploit the extra degrees of freedom (DoFs) by allowing individual rotation and scaling for each constellation point. By jointly optimizing the transmitting signal matrix and the complex scaling factors, the signal-to-interference-and-noise-ratio (SINR) at the receiver is maximized. The formulated optimization problem is divided into two sub-problems and solved by the alternating optimization approach. We further propose a simplified gradient descent algorithm and extend our algorithm to frame-level precoding design. Simulation results show that the proposed algorithms outperform the existing CEP algorithms in terms of both symbol error rate (SER) and computational complexity.

Distributed Algorithms for Spectral and Energy-Efficiency Maximization of K-User Interference Channels

In this paper, we propose a cooperative distributed framework to optimize a variety of rate and energy-efficiency (EE) utility functions, such as the minimum-weighted rate or the global EE, for the K-user interference channel. We focus on the single-input multiple-output (SIMO) case, where each user, based solely on local channel state information (CSI) and limited exchange information from other users, optimizes its transmit power and receive beamformer, although the framework can also be extended to the multiple-output multiple-input (MIMO) case. The distributed framework combines an alternating optimization approach with majorization-minimization (MM) techniques, thus ensuring convergence to a stationary point of the centralized cost function. Closed-form power update rules are obtained for some utility functions, thus obtaining very fast convergence algorithms. The receivers treat interference as noise (TIN) and apply the beamformers that maximize the signal-to-interference-plus-noise (SINR). The proposed cooperative distributed algorithms are robust against channel variations and network topology changes and, as our simulation results suggest, they perform close to the centralized solution that requires global CSI. As a benchmark, we also study a non-cooperative distributed framework based on the so-called “signal-to-leakage-plus-noise ratio” (SNLR) that further reduces the overhead of the cooperative version.

Design optimization of a hybrid solar-biomass plant to sustainably supply energy to industry: Methodology and case study

To gain the benefits of solar-biomass cogeneration plant, supplying energy to industry, the design optimization procedure is required to holistically integrate economical, technical and environmental aspects. This entails formulation of a performance criterion that maximizes solar fraction, reduces investment cost, lowers thermal storage loss and puts less pressure on biomass resources. It is also necessary to consider factors influencing the plant’s performance while giving active role to industrial demand. This kind of optimization approach adaptively evolves as influences change and is not static. However, such a criterion is not yet part in many of the existing design optimization schemes. In this work, an alternative optimization approach that addresses the aforementioned issues is proposed. To this end, a molten salt biomass boiler is modeled in MATLAB and integrated to a solar plant model in TRNSYS. The resulting configuration is latter optimized in GenOpt to minimize biomass power utilization index (BPUI) and excess saturated steam generation (ESSG). Demonstrated by a case study, a plant efficiency of 31.5% with 23.5% solar gain is optimally designed resulting in about 0.094$/kWh levelized cost of generation. Furthermore, considering global power outage loss, the hybrid plant could be seen as a preferred industrial source of energy.

Service operation design in a transit network with congested common lines

This paper focuses on a transit service operation design problem that primarily determines the optimal frequency settings with explicit consideration of congested common lines in a bus service network. Other than passengers’ transit route choices, the transit service line choices among congested common lines are also specified in the model formulation. A tri-level programming approach is applied to formulate this problem, wherein the upper-level program optimizes the transit frequency to minimize the total operating costs and passengers’ transit costs; the middle-level program describes passengers’ transit routing choices, in which passengers will select a sequence of transfer nodes to minimize their transit costs; and the lower-level program formulates the equilibrium strategy in the common line problem on the route sections (i.e., between two successive transfer nodes), whose equilibrium solution may have multiple strategies depending on the congestion level of the common lines. The tri-level model is then reformulated into a mathematical program with equilibrium constraints. Two solution methods are proposed to solve the problem. One is to transform the model into a mixed-integer linear program so that the global optimal solution of the linearized problem can be guaranteed, and the other employs a surrogate optimization approach to ensure high solution efficiency for large size problems without compromising solution quality. Finally, we conduct extensive numerical examples to demonstrate the validity of our model formulation and the performance of the proposed solution algorithms.

Deep Convolutional Neural Network Compression via Coupled Tensor Decomposition

Large neural networks have aroused impressive progress in various real world applications. However, the expensive storage and computational resources requirement for running deep networks make them problematic to be deployed on mobile devices. Recently, matrix and tensor decompositions have been employed for compressing neural networks. In this paper, we develop a simultaneous tensor decomposition technique for network optimization. The shared network structure is first discussed. Sometimes, not only the structure but also the parameters are shared to form a compressed model at the expense of degraded performance. This indicates that the weight tensors between layers within one network contain both identical components and independent components. To utilize this characteristic, two new coupled tensor train decompositions are developed for fully and partly structure sharing cases, and an alternating optimization approach is proposed for low rank tensor computation. Finally, we restore the performance of the neural network model by fine-tuning. The compression ratio of the devised approach can then be calculated. Experimental results are also included to demonstrate the benefits of our algorithm for both applications of image reconstruction and classification, using the well known datasets such as Cifar-10/Cifar-100 and ImageNet and widely used networks such as ResNet. Comparing to the state-of-the-art independent matrix and tensor decomposition based methods, our model can obtain a better network performance under the same compression ratio.

Pareto-Optimal Pilot Design for Cellular Massive MIMO Systems

We introduce a non-orthogonal pilot design scheme that simultaneously minimizes two contradicting targets of channel estimation errors of all base stations (BSs) and the total pilot power consumption of all users in a multi-cell massive MIMO system, subject to the transmit power constraints of the users in the network. We formulate a multi-objective optimization problem (MOP) with two objective functions capturing the contradicting targets and find the Pareto optimal solutions for the pilot signals. Using weighted-sum-scalarization technique, we first convert the MOP to an equivalent single-objective optimization problem (SOP), which is not convex. Assuming that each BS is provided with the most recent knowledge of the pilot signals of the other BSs, we then decompose the SOP into a set of distributed non-convex optimization problems to be solved at individual BSs. Finally, we introduce an alternating optimization approach to cast each one of the resulting distributed optimization problems into a convex linear matrix inequality (LMI) form. We provide a mathematical proof for the convergence of the proposed alternating approach and a complexity analysis for the LMI optimization problem. Simulation results confirm that the proposed approach can significantly reduce pilot power, whilst maintaining the same level of channel estimation error as in a baseline.

Curved fibre path optimisation for improved shape adaptive composite propeller blade design

The propulsive efficiency of a shape adaptive composite propeller increases with passive pitch (twist) change under a hydrodynamic load. These adaptive marine propellers are usually designed using straight fibre tows to obtain optimised layup with such bend-twist character. This paper aims to present an alternative optimisation approach for such composite propellers using a curved fibre path method where tow paths are allowed to vary spatially within individual plies. In this work, three optimisation studies were performed to highlight that employing curved tows produce better results, including increased twist change for efficiency or, lowered deflection for structural improvements. Laminate plates were first optimised and manufactured using an automated fibre placement robot for experimental validation. Cantilevered loading test results show that optimised curved fibre plates generated >10% more twist compared to their straight fibre counterparts. In the following optimisation on a hydrofoil model, the curved fibre design produced the same tip angle change as the optimised straight fibre layup, yet it reduced the tip deflection by 15%. Finally, the optimisation was performed on a Wageningen B-series marine propeller, in which the optimised blade design with curved fibre layup achieved a 20% reduction in the Tsai-Hill failure index under the same pitch change.

Energy Management and Trajectory Optimization for UAV-Enabled Legitimate Monitoring Systems

Thanks to their quick placement and high flexibility, unmanned aerial vehicles (UAVs) can be very useful in the current and future wireless communication systems. With a growing number of smart devices and infrastructure-free communication networks, it is necessary to legitimately monitor these networks to prevent crimes. In this paper, a novel framework is proposed to exploit the flexibility of the UAV for legitimate monitoring via joint trajectory design and energy management. The system includes a suspicious transmission link with a terrestrial transmitter and a terrestrial receiver, and a UAV to monitor the suspicious link. The UAV can adjust its positions and send jamming signal to the suspicious receiver to ensure successful eavesdropping. Based on this model, we first develop an approach to minimize the overall jamming energy consumption of the UAV. Building on a judicious (re-)formulation, an alternating optimization approach is developed to compute a locally optimal solution in polynomial time. Furthermore, we model and include the propulsion power to minimize the overall energy consumption of the UAV. Leveraging the successive convex approximation method, an effective iterative approach is developed to find a feasible solution fulfilling the Karush-Kuhn-Tucker (KKT) conditions. Extensive numerical results are provided to verify the merits of the proposed schemes.

Robust Probabilistic-Constrained Optimization for IRS-Aided MISO Communication Systems

Taking into account imperfect channel state information, this letter formulates and solves a joint active/passive beamforming optimization problem in multiple-input single-output systems with the support of an intelligent reflecting surface. In particular, we introduce an optimization problem to minimize the total transmit power subject to maintaining the users' signal-to-interference-plus-noise-ratio coverage probability above a predefined target. Due to the presence of probabilistic constraints, the proposed optimization problem is non-convex. To circumvent this issue, we first recast the proposed problem in a convex form by adopting the Bernstein-type inequality, and we then introduce a converging alternating optimization approach to iteratively find the active/passive beamforming vectors. In particular, the transformed robust optimization problem can be effectively solved by using standard interior-point methods. Numerical results demonstrate the effectiveness of jointly optimizing the active/passive beamforming vectors.

Full-duplex NOMA cellular networks: Beamforming design and user scheduling

We consider a full-duplex (FD) cellular network, where two pairs of downlink and uplink users simultaneously communicate with a multi-antenna FD base station (BS) via non-orthogonal multiple access (NOMA). The downlink sum-rate of the system is maximized by optimizing the receive and transmit beamformers at the BS, while specific rates for the uplink users are guaranteed. The resulting non-convex problem is solved using an alternating optimization approach based on semi-definite relaxation and one-dimensional line-search approach. To achieve a low-complexity implementation, we incorporate zero-forcing self-interference suppression constraint at the BS, and derive closed-form expression for the receive beamformer. In order to manage the interference between the downlink and uplink users, two user scheduling algorithms, namely Nearest Near and Furthest Far user (NNFF) Scheduling and Random Near and Random Far (RNRF) User Scheduling are proposed based on the cell partitioning method. Numerical results show that FD-NOMA system with optimum and suboptimum beamforming designs significantly improve the system’s ergodic sum-rate compared to the half-duplex counterpart. Moreover, in the FD-NOMA system, NNFF user scheduling can achieve up to 25% average sum-rate gain as compared with the RNRF user scheduling algorithm.

Automatic feature scaling and selection for support vector machine classification with functional data

FunctionalData Analysis (FDA) has become a very important field in recent years due to its wide range of applications. However, there are several real-life applications in which hybrid functional data appear, i.e., data with functional and static covariates. The classification of such hybrid functional data is a challenging problem that can be handled with the Support Vector Machine (SVM). Moreover, the selection of the most informative features may yield to drastic improvements in the classification rates. In this paper, an embedded feature selection approach for SVM classification is proposed, in which the isotropic Gaussian kernel is modified by associating a bandwidth to each feature. The bandwidths are jointly optimized with the SVM parameters, yielding an alternating optimization approach. The effectiveness of our methodology was tested on benchmark data sets. Indeed, the proposed method achieved the best average performance when compared to 17 other feature selection and SVM classification approaches. A comprehensive sensitivity analysis of the parameters related to our proposal was also included, confirming its robustness.

Joint Active and Passive Beamforming Optimization for Intelligent Reflecting Surface Assisted SWIPT Under QoS Constraints

Intelligent reflecting surface (IRS) is a new and revolutionizing technology for achieving spectrum and energy efficient wireless networks. By leveraging massive low-cost passive elements that are able to reflect radio-frequency (RF) signals with adjustable phase shifts, IRS can achieve high passive beamforming gains, which are particularly appealing for improving the efficiency of RF-based wireless power transfer. Motivated by the above, we study in this paper an IRS-assisted simultaneous wireless information and power transfer (SWIPT) system. Specifically, a set of IRSs are deployed to assist in the information/power transfer from a multi-antenna access point (AP) to multiple single-antenna information users (IUs) and energy users (EUs), respectively. We aim to minimize the transmit power at the AP via jointly optimizing its transmit precoders and the reflect phase shifts at all IRSs, subject to the quality-of-service (QoS) constraints at all users, namely, the individual signal-to-interference-plus-noise ratio (SINR) constraints at IUs and the energy harvesting constraints at EUs. However, this optimization problem is non-convex with intricately coupled variables, for which the existing alternating optimization approach is shown to be inefficient as the number of QoS constraints increases. To tackle this challenge, we first apply proper transformations on the QoS constraints and then propose an efficient iterative algorithm by applying the penalty-based optimization method. Moreover, by exploiting the short-range coverage of IRS, we further propose a more computationally efficient algorithm by optimizing the phase shifts at all IRSs in parallel. Simulation results demonstrate the effectiveness of employing multiple IRSs for enhancing the performance of SWIPT systems as well as the significant performance gains achieved by our proposed algorithms over benchmark schemes. The impact of IRS on the transmitter/receiver design for SWIPT is also unveiled.

A biological effect‐guided optimization approach using beam distal‐edge avoidance for intensity‐modulated proton therapy

Linear energy transfer (LET)-guided methods have been applied to intensity-modulated proton therapy (IMPT) to improve its biological effect. However, using LET as a surrogate for biological effect ignores the topological relationship of the scanning spot to different structures of interest. In this study, we developed an optimization method that takes advantage of the continuing increase in LET beyond the physical dose Bragg peak. This method avoids placing high biological effect values in critical structures and increases biological effect in the tumor area without compromising target coverage. We selected the cases of two patients with brain tumors and two patients with head and neck tumors who had been treated with proton therapy at our institution. Three plans were created for each case: a plan based on conventional dose-based optimization (DoseOpt), one based on LET-incorporating optimization (LETOpt), and one based on the proposed distal-edge avoidance-guided optimization method (DEAOpt). In DEAOpt, an L1 -norm sparsity term, in which the penalty of each scanning spot was set according to the topological relationship between the organ positions and the location of the peak scaled LET-weighted dose (c LETxD) was added to a conventional dose-based optimization objective function. All plans were normalized to give the same target dose coverage. Dose (assuming a constant relative biological effectiveness value of 1.1, as in clinical practice), biological effect (c LETxD), and computing time consumption were evaluated and compared among the three optimization approaches for each patient case. For all four cases, all three optimization methods generated comparable dose coverage in both target and critical structures. The LETOpt plans and DEAOpt plans reduced biological effect hot spots in critical structures and increased biological effect in the target volumes to a similar extent. For the target, the c LETxD98% and c LETxD2% in the DEAOpt plans were on average 7.2% and 11.74% higher than in the DoseOpt plans, respectively. For the brainstem, the c LETxDmean in the DEAOpt plans was on average 33.38% lower than in the DoseOpt plans. In addition, the DEAOpt method saved 30.37% of the computation cost over the LETOpt method. DEAOpt is an alternative IMPT optimization approach that correlates the location of scanning spots with biological effect distribution. IMPT could benefit from the use of DEAOpt because this method not only delivers comparable biological effects to LETOpt plans, but also is faster.

Robust Energy Efficiency Optimization Algorithm for Health Monitoring System With Wireless Body Area Networks

The electronic health monitoring system, with wireless body area networks (WBANs), is a new technological paradigm that helps early detection of any abnormal physiological signs in a patient’s body. In this letter, an optimization problem is formulated to optimize the energy efficiency (EE) of WBAN without requiring the channel state information (CSI) from the transmitting sensor nodes to the aggregator. We utilize a generalized gamma distribution that supports various patient conditions during daily life activities and can efficiently model both everyday and dynamic activities. The optimization problem aims to optimize each sensor transmit power and encoding rate to minimize the EE (measured in J/bits) by considering outage probability and packet retransmission. It is shown that the formulated optimization problem is semi-strictly quasi-convex in each decision variable, and an alternative optimization approach is proposed to determine its solution. The simulation results show that the proposed algorithm is 30% more energy efficient when compared to sub-optimal solution with a constant encoding rate and transmit power of 5 dB, and 0.6 mW, respectively.