Research Outputs
Permanent URI for this communityhttps://hdl.handle.net/20.500.14288/2
Browse
10 results
Search Results
Publication Metadata only Antenna array structures for enhanced cluster index modulation(IEEE, 2023) Koc, Asil; Le-Ngoc, Tho; Department of Electrical and Electronics Engineering; Raeisi, Mahmoud; Yıldırım, İbrahim; Başar, Ertuğrul; Department of Electrical and Electronics Engineering; Graduate School of Sciences and Engineering; College of Engineering; Communications Research and Innovation Laboratory (CoreLab)This paper investigates the effect of various antenna array structures, i.e., uniform linear array (ULA), uniform rectangular array (URA), uniform circular array (UCA), and concentric circular array (CCA), on cluster index modulation (CIM) enabled massive multiple-input multiple-output (mMIMO) millimeter-wave (mmWave) communications systems. As the CIM technique indexes spatial clusters to convey additional information bits, the different radiation characteristics caused by different array structures can significantly affect system performance. By analyzing the effects of array characteristics such as radiation pattern, array directivity, half-power beam width (HPBW), and radiation side lobes on bit error rate (BER) performance, we reveal that URA achieves better error performance than its counterparts in a CIM-enabled mmWave system. We demonstrate that narrower beams alone cannot guarantee better BER performance in a CIM-based system. Instead, other radiation characteristics, especially radiation side lobes, can significantly influence system performance by entailing extra interference in the non-intended directions. Illustrative results show that URA owes its superiority to its lower side lobes. We also propose an algorithm to implement fixed phase shifters (FPS) as a hardware-efficient (HE) analog network structure (beamformer/combiner) to reduce cost and energy consumption in mmWave systems and investigate the effect of a non-ideal analog network on the BER performance for different array structures. It is demonstrated that HE systems with a few FPSs can achieve similar BER performance compared to the optimum (OP) analog network structure.Publication Metadata only Detection and mitigation of targeted data poisoning attacks in federated learning(IEEE, 2022) Department of Computer Engineering; Erbil, Pınar; Gürsoy, Mehmet Emre; Department of Computer Engineering; College of EngineeringFederated learning (FL) has emerged as a promising paradigm for distributed training of machine learning models. In FL, several participants train a global model collaboratively by only sharing model parameter updates while keeping their training data local. However, FL was recently shown to be vulnerable to data poisoning attacks, in which malicious participants send parameter updates derived from poisoned training data. In this paper, we focus on defending against targeted data poisoning attacks, where the attacker's goal is to make the model misbehave for a small subset of classes while the rest of the model is relatively unaffected. To defend against such attacks, we first propose a method called MAPPS for separating malicious updates from benign ones. Using MAPPS, we propose three methods for attack detection: MAPPS + X-Means, MAPPS + VAT, and their Ensemble. Then, we propose an attack mitigation approach in which a "clean" model (i.e., a model that is not negatively impacted by an attack) can be trained despite the existence of a poisoning attempt. We empirically evaluate all of our methods using popular image classification datasets. Results show that we can achieve > 95% true positive rates while incurring only < 2% false positive rate. Furthermore, the clean models that are trained using our proposed methods have accuracy comparable to models trained in an attack-free scenario.Publication Metadata only Histopathological classification of colon tissue images with self-supervised models(IEEE, 2023) Department of Computer Engineering; Erden, Mehmet Bahadır; Cansız, Selahattin; Demir, Çiğdem Gündüz; Department of Computer Engineering; Graduate School of Sciences and Engineering; College of EngineeringDeep learning techniques have demonstrated their ability to facilitate medical image diagnostics by offering more precise and accurate predictions. Convolutional neural network (CNN) architectures have been employed for a decade as the primary approach to enable automated diagnosis. On the other hand, recently proposed vision transformers (ViTs) based architectures have shown success in various computer vision tasks. However, their efficacy in medical image classification tasks remains largely unexplored due to their requirement for large datasets. Nevertheless, significant performance gains can be achieved by leveraging self-supervised learning techniques through pretraining. This paper analyzes performance of self-supervised pretrained networks in medical image classification tasks. Results on colon histopathology images revealed that CNN based architectures are more effective when trained from scratch, while pretrained models could achieve similar levels of performance with limited data.Publication Metadata only Implications of node selection in decentralized federated learning(IEEE, 2023) Department of Computer Engineering; Lodhi, Ahnaf Hannan; Akgün, Barış; Özkasap, Öznur; Department of Computer Engineering; Graduate School of Sciences and Engineering; College of EngineeringDecentralized Federated Learning (DFL) offers a fully distributed alternative to Federated Learning (FL). However, the lack of global information in a highly heterogeneous environment negatively impacts its performance. Node selection in FL has been suggested to improve both communication efficiency and convergence rate. In order to assess its impact on DFL performance, this work evaluates node selection using performance metrics. It also proposes and evaluates a time-varying parameterized node selection method for DFL employing validation accuracy and its per-round change. The mentioned criteria are evaluated using both hard and stochastic/soft selection on sparse networks. The results indicate that the bias associated with node selection adversely impacts performance as training progresses. Furthermore, under extreme conditions, soft selection is observed to result in higher diversity for better generalization, while a completely random selection is shown to be preferable with very limited participation.Publication Metadata only Measurement and characteristic analysis of ris-assisted wireless communication channels in sub-6 ghz outdoor scenarios(IEEE, 2023) Lan, Jifeng; Sang, Jian; Zhou, Mingyong; Gao, Boning; Meng, Shengguo; Li, Xiao; Tang, Wankai; Jin, Shi; Cheng, Qiang; Cuit, Tie Jun; Department of Electrical and Electronics Engineering; Başar, Ertuğrul; Department of Electrical and Electronics Engineering; College of EngineeringReconfigurable intelligent surface (RIS)-empowered communication has recently drawn significant attention due to its superior capability in manipulating the wireless propagation environment. However, the channel modeling and measurement of RIS-assisted wireless communication systems in real environment has not been adequately studied. In this paper, we construct a channel measurement system using vector network analyzer (VNA) is used to investigate RIS-assisted wireless communication channel in outdoor scenarios at 2.6 GHz. New path loss (PL) models including angle domain information are proposed by refining the traditional close-in (CI) and floating-intercept (FI) models. In the proposed models, both influences of the distance from transmitter (TX) to RIS and the distance from receiver (RX) to RIS on the PL, are taken into account. In addition, the value of root mean square (RMS) delay spread of RIS-assisted wireless communication is found to be much smaller than that of the traditional non line-of-sight (NLOS) case, implying that RIS provides a virtual line-of-sight (LOS) link.Publication Metadata only Measurement-based characterization of physical layer security for ris-assisted wireless systems(IEEE, 2023) Kesir, Samed; Wikelek, Ibrahim; Pusane, Ali Emre; Gorcin, Ali; Department of Electrical and Electronics Engineering; Kayraklık, Sefa; Başar, Ertuğrul; Department of Electrical and Electronics Engineering; CoreLab; Graduate School of Sciences and Engineering; College of EngineeringThere have been recently many studies demonstrating that the performance of wireless communication systems can be significantly improved by a reconfigurable intelligent surface (RIS), which is an attractive technology due to its low power requirement and low complexity. This paper presents a measurement-based characterization of RISs for providing physical layer security, where the transmitter (Alice), the intended user (Bob), and the eavesdropper (Eve) are deployed in an indoor environment. Each user is equipped with a software-defined radio connected to a horn antenna. The phase shifts of reflecting elements are software controlled to collaboratively determine the amount of received signal power at the locations of Bob and Eve in such a way that the secrecy capacity is aimed to be maximized. An iterative method is utilized to configure a Greenerwave RIS prototype consisting of 76 passive reflecting elements. Computer simulation and measurement results demonstrate that an RIS can be an effective tool to significantly increase the secrecy capacity between Bob and Eve.Publication Metadata only Physical activity recognition using deep transfer learning with convolutional neural networks(IEEE, 2022) Department of Electrical and Electronics Engineering;Department of Computer Engineering; Ataseven, Berke; Madani, Alireza; Gürsoy, Beren Semiz; Gürsoy, Mehmet Emre; Graduate School of Sciences and Engineering; College of EngineeringCurrent wearable devices are capable of monitoring various health indicators as well as fitness and/or physical activity types. However, even on the latest models of many wearable devices, users need to manually enter the type of work-out or physical activity they are performing. In order to automate real-time physical activity recognition, in this study, we develop a deep transfer learning-based physical activity recognition framework using acceleration data acquired through inertial measurement units (IMUs). Towards this goal, we modify a pre-trained version of the GoogLeNet convolutional neural network and fine-tune it with data from IMUs. To make IMU data compatible with GoogLeNet, we propose three novel data transform approaches based on continuous wavelet transform: Horizontal Concatenation (HC), Acceleration-Magnitude (AM), and Pixelwise Axes-Averaging (PA). We evaluate the performance of our approaches using the real-world PAMAP2 dataset. The three approaches result in 0.93, 0.95 and 0.98 validation accuracy and 0.75, 0.85 and 0.91 test accuracy, respectively. The PA approach yields the highest weighted F1 score (0.91) and activity-specific true positive ratios. Overall, our methods and results show that accurate real-time physical activity recognition can be achieved using transfer learning and convolutional neural networks.Publication Metadata only Role of audio in video summarization(IEEE, 2023) Department of Computer Engineering; Shoer, İbrahim; Köprü, Berkay; Erzin, Engin; Department of Computer Engineering; Koç Üniversitesi İş Bankası Yapay Zeka Uygulama ve Araştırma Merkezi (KUIS AI)/ Koç University İş Bank Artificial Intelligence Center (KUIS AI); Graduate School of Sciences and Engineering; College of EngineeringVideo summarization attracts attention for efficient video representation, retrieval, and browsing to ease volume and traffic surge problems. Although video summarization mostly uses the visual channel for compaction, the benefits of audio-visual modeling appeared in recent literature. The information coming from the audio channel can be a result of audio-visual correlation in the video content. In this study, we propose a new audio-visual video summarization framework integrating four ways of audio-visual information fusion with GRU-based and attention-based networks. Furthermore, we investigate a new explainability methodology using audio-visual canonical correlation analysis (CCA) to better understand and explain the role of audio in the video summarization task. Experimental evaluations on the TVSum dataset attain F1 score and Kendall-tau score improvements for the audio-visual video summarization. Furthermore, splitting video content on TVSum and COGNIMUSE datasets based on audio-visual CCA as positively and negatively correlated videos yields a strong performance improvement over the positively correlated videos for audio-only and audio-visual video summarization.Publication Metadata only Secure hierarchical federated learning in vehicular networks using dynamic client selection and anomaly detection(IEEE-Inst Electrical Electronics Engineers Inc, 2024) ; Department of Electrical and Electronics Engineering; Haghighifard, Mohammad Saeid; Ergen, Sinem Çöleri; Department of Electrical and Electronics Engineering; ; Graduate School of Sciences and Engineering; College of Engineering;Hierarchical Federated Learning (HFL) faces the significant challenge of adversarial or unreliable vehicles in vehicular networks, which can compromise the model's integrity through misleading updates. Addressing this, our study introduces a novel framework that integrates dynamic vehicle selection and robust anomaly detection mechanisms, aiming to optimize participant selection and mitigate risks associated with malicious contributions. Our approach involves a comprehensive vehicle reliability assessment, considering historical accuracy, contribution frequency, and anomaly records. An anomaly detection algorithm is utilized to identify anomalous behavior by analyzing the cosine similarity of local or model parameters during the federated learning (FL) process. These anomaly records are then registered and combined with past performance for accuracy and contribution frequency to identify the most suitable vehicles for each learning round. Dynamic client selection and anomaly detection algorithms are deployed at different levels, including cluster heads (CHs), cluster members (CMs), and the Evolving Packet Core (EPC), to detect and filter out spurious updates. Through simulation-based performance evaluation, our proposed algorithm demonstrates remarkable resilience even under intense attack conditions. Even in the worst-case scenarios, it achieves convergence times at 63 % as effective as those in scenarios without any attacks. Conversely, in scenarios without utilizing our proposed algorithm, there is a high likelihood of non-convergence in the FL process. © 2024 IEEE.Publication Metadata only Uncovering public attitudes: utilizing a PCA-based approach to examine anti-immigrant attitudes(Institute of Electrical and Electronics Engineers Inc., 2024) Öktem, Ubeyd; Graduate School of Social Sciences and HumanitiesThis research will observe the determinants of public attitudes towards immigrants in Turkey depending on survey questions from 2016 and 2019. The literature suggests the effects of several variables on the perception towards immigrants, with specific emphasis on encounter level and economic conditions. This research aims to test these theories in Turkey's case with migration and observe their effects in time. Although the primary intention is to examine these relations, the paper also claims to utilize unsupervised machine learning methods in variable construction/indexing. Therefore, the other aspect of this research is to utilize Principal Component Analysis in variable construction as opposed to equally weighting different variables. As a combination of quantitative and computational sociology methodologies, this paper will initially answer two forms of questions: one sociological and the other methodological. © 2024 IEEE.