Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081365
K. Tziridis, T. Kalampokas, G. Papakostas, K. Diamantaras
This paper deals with the problem of airfare prices prediction. For this purpose a set of features characterizing a typical flight is decided, supposing that these features affect the price of an air ticket. The features are applied to eight state of the art machine learning (ML) models, used to predict the air tickets prices, and the performance of the models is compared to each other. Along with the prediction accuracy of each model, this paper studies the dependency of the accuracy on the feature set used to represent an airfare. For the experiments a novel dataset consisting of 1814 data flights of the Aegean Airlines for a specific international destination (from Thessaloniki to Stuttgart) is constructed and used to train each ML model. The derived experimental results reveal that the ML models are able to handle this regression problem with almost 88% accuracy, for a certain type of flight features.
{"title":"Airfare prices prediction using machine learning techniques","authors":"K. Tziridis, T. Kalampokas, G. Papakostas, K. Diamantaras","doi":"10.23919/EUSIPCO.2017.8081365","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081365","url":null,"abstract":"This paper deals with the problem of airfare prices prediction. For this purpose a set of features characterizing a typical flight is decided, supposing that these features affect the price of an air ticket. The features are applied to eight state of the art machine learning (ML) models, used to predict the air tickets prices, and the performance of the models is compared to each other. Along with the prediction accuracy of each model, this paper studies the dependency of the accuracy on the feature set used to represent an airfare. For the experiments a novel dataset consisting of 1814 data flights of the Aegean Airlines for a specific international destination (from Thessaloniki to Stuttgart) is constructed and used to train each ML model. The derived experimental results reveal that the ML models are able to handle this regression problem with almost 88% accuracy, for a certain type of flight features.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"240 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122996787","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081471
D. Varga, T. Szirányi
Person re-identification is one of the widely studied research topic in the fields of computer vision and pattern recognition. In this paper, we present a deep multi-instance learning approach for person re-identification. Since most publicly available databases for pedestrian re-identification are not enough big, over-fitting problems occur in deep learning architectures. To tackle this problem, person re-identification is expressed as a deep multi-instance learning issue. Therefore, a multi-scale feature learning process is introduced which is driven by optimizing a novel cost function. We report on experiments and comparisons to other state-of-the-art algorithms using publicly available databases such as VIPeR and ETHZ.
{"title":"Person re-identification based on deep multi-instance learning","authors":"D. Varga, T. Szirányi","doi":"10.23919/EUSIPCO.2017.8081471","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081471","url":null,"abstract":"Person re-identification is one of the widely studied research topic in the fields of computer vision and pattern recognition. In this paper, we present a deep multi-instance learning approach for person re-identification. Since most publicly available databases for pedestrian re-identification are not enough big, over-fitting problems occur in deep learning architectures. To tackle this problem, person re-identification is expressed as a deep multi-instance learning issue. Therefore, a multi-scale feature learning process is introduced which is driven by optimizing a novel cost function. We report on experiments and comparisons to other state-of-the-art algorithms using publicly available databases such as VIPeR and ETHZ.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128275367","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081607
Mario Castanheira, Antônio Simoes, M. Gomes, R. Dinis, V. Silva
This paper proposes a transmitter structure that combines a ring-type magnitude modulation (RMM) technique with a linear amplification with nonlinear components (LINC) scheme for power and spectrally efficient transmission based on bandwidth limited OQPSK signals, for either a linear combiner (LC) or a Chireix combiner (CC). It shows that by controlling the transmitted signal's envelope through RMM, the range of the LINC decomposition angle is considerably decreased. This significantly improves LC's power efficiency, and substantially reduces CC's spectral leakage while maintaining its high amplification efficiency.
{"title":"Boosting LINC systems combiners efficiency through ring-type magnitude modulation","authors":"Mario Castanheira, Antônio Simoes, M. Gomes, R. Dinis, V. Silva","doi":"10.23919/EUSIPCO.2017.8081607","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081607","url":null,"abstract":"This paper proposes a transmitter structure that combines a ring-type magnitude modulation (RMM) technique with a linear amplification with nonlinear components (LINC) scheme for power and spectrally efficient transmission based on bandwidth limited OQPSK signals, for either a linear combiner (LC) or a Chireix combiner (CC). It shows that by controlling the transmitted signal's envelope through RMM, the range of the LINC decomposition angle is considerably decreased. This significantly improves LC's power efficiency, and substantially reduces CC's spectral leakage while maintaining its high amplification efficiency.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124644474","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081312
N. Nikolaidis, Charalambos Symeonidis
Most activity-based person identity recognition methods operate on video data. Moreover, the vast majority of these methods focus on gait recognition. Obviously, recognition of a subject's identity using only gait imposes limitations to the applicability of the corresponding methods whereas a method capable of recognizing the subject's identity from various activities would be much more widely applicable. In this paper, a new method for activity-based identity recognition operating on motion capture data, that can recognize the subject's identity from a variety of activities is proposed. The method combines an existing approach for feature extraction from motion capture sequences with a label propagation algorithm for classification. The method and its variants (including a novel one, that takes advantage of the fact that, in certain cases, both activity and person identity labels might exist for the labeled sequences) have been tested in two different datasets. Experimental analysis proves that the proposed approach provides very good person identity recognition results, surpassing those obtained by two other methods.
{"title":"Person identity recognition on motion capture data using label propagation","authors":"N. Nikolaidis, Charalambos Symeonidis","doi":"10.23919/EUSIPCO.2017.8081312","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081312","url":null,"abstract":"Most activity-based person identity recognition methods operate on video data. Moreover, the vast majority of these methods focus on gait recognition. Obviously, recognition of a subject's identity using only gait imposes limitations to the applicability of the corresponding methods whereas a method capable of recognizing the subject's identity from various activities would be much more widely applicable. In this paper, a new method for activity-based identity recognition operating on motion capture data, that can recognize the subject's identity from a variety of activities is proposed. The method combines an existing approach for feature extraction from motion capture sequences with a label propagation algorithm for classification. The method and its variants (including a novel one, that takes advantage of the fact that, in certain cases, both activity and person identity labels might exist for the labeled sequences) have been tested in two different datasets. Experimental analysis proves that the proposed approach provides very good person identity recognition results, surpassing those obtained by two other methods.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124666901","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081646
R. Takashima, Y. Kawaguchi, M. Togami
This paper proposes a sound source separation method for vibration-derived sound signals such as sounds derived from mechanical vibrations by using vibration sensors. The proposed method is based on two assumptions. First, a vibration signal and the sound derived from the vibration are assumed to have a linear correlation. This assumption enables us to model the vibration-derived sound as a linear convolution of a transfer function and a vibration signal recorded by a vibration sensor. Second, un-vibration-derived sound signals such that the sound source is not connected to vibration sensors via a solid medium are barely recorded by vibration sensors. This assumption leads to a constraint of the transfer function from the un-vibration-derived sound sources to the vibration sensors. The proposed framework is the same as a microphone-array-based blind source separation framework, except that the proposed method constructs arrays with microphones and vibration sensors, and the separation parameters are constrained by the prior knowledge gained from the above second assumption. Experimental results indicate that the separation performance of the proposed method is superior to that of a conventional microphone-array-based source separation method.
{"title":"Separation of vibration-derived sound signals based on fusion processing of vibration sensors and microphones","authors":"R. Takashima, Y. Kawaguchi, M. Togami","doi":"10.23919/EUSIPCO.2017.8081646","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081646","url":null,"abstract":"This paper proposes a sound source separation method for vibration-derived sound signals such as sounds derived from mechanical vibrations by using vibration sensors. The proposed method is based on two assumptions. First, a vibration signal and the sound derived from the vibration are assumed to have a linear correlation. This assumption enables us to model the vibration-derived sound as a linear convolution of a transfer function and a vibration signal recorded by a vibration sensor. Second, un-vibration-derived sound signals such that the sound source is not connected to vibration sensors via a solid medium are barely recorded by vibration sensors. This assumption leads to a constraint of the transfer function from the un-vibration-derived sound sources to the vibration sensors. The proposed framework is the same as a microphone-array-based blind source separation framework, except that the proposed method constructs arrays with microphones and vibration sensors, and the separation parameters are constrained by the prior knowledge gained from the above second assumption. Experimental results indicate that the separation performance of the proposed method is superior to that of a conventional microphone-array-based source separation method.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130472903","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081510
Anshul Thakur, R. Jyothi, Padmanabhan Rajan, A. D. Dileep
Bird activity detection is the task of determining if a bird sound is present in a given audio recording. This paper describes a bird activity detector which utilises a support vector machine (SVM) with a dynamic kernel. Dynamic kernels are used to process sets of feature vectors having different cardinalities. Probabilistic sequence kernel (PSK) is one such dynamic kernel. The PSK converts a set of feature vectors from a recording into a fixed-length vector. We propose to use a variant of PSK in this work. Before computing the fixed-length vector, cepstral mean and variance normalisation and short-time Gaussianization is performed on the feature vectors. This reduces environment mismatch between different recordings. Additionally, we also demonstrate a simple procedure to speed up the proposed method by reducing the size of fixed-length vector. A speedup of almost 70% is observed, with a very small drop in accuracy. The proposed method is also compared with a random forest classifier and is shown to outperform it.
{"title":"Rapid bird activity detection using probabilistic sequence kernels","authors":"Anshul Thakur, R. Jyothi, Padmanabhan Rajan, A. D. Dileep","doi":"10.23919/EUSIPCO.2017.8081510","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081510","url":null,"abstract":"Bird activity detection is the task of determining if a bird sound is present in a given audio recording. This paper describes a bird activity detector which utilises a support vector machine (SVM) with a dynamic kernel. Dynamic kernels are used to process sets of feature vectors having different cardinalities. Probabilistic sequence kernel (PSK) is one such dynamic kernel. The PSK converts a set of feature vectors from a recording into a fixed-length vector. We propose to use a variant of PSK in this work. Before computing the fixed-length vector, cepstral mean and variance normalisation and short-time Gaussianization is performed on the feature vectors. This reduces environment mismatch between different recordings. Additionally, we also demonstrate a simple procedure to speed up the proposed method by reducing the size of fixed-length vector. A speedup of almost 70% is observed, with a very small drop in accuracy. The proposed method is also compared with a random forest classifier and is shown to outperform it.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126843478","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081224
Ryohei Sasaki, K. Konishi, Tomohiro Takahashi, T. Furukawa
This paper proposes a new algorithm for image inpainting algorithm based on the matrix rank minimization with nonlinear mapping function. Assuming that each intensity value of a nonlinear mapped image can be modeled by the autoregressive (AR) model, the image inpainting problem is formulated as a kind of the matrix rank minimization problem, and this paper modifies the iterative partial matrix shrinkage (IPMS) algorithm and provides an inpainting algorithm, which estimates a nonlinear mapping function and the missing pixels simultaneously. Numerical examples show that the proposed algorithm recovers missing pixels efficiently.
{"title":"Low-rank and nonlinear model approach to image inpainting","authors":"Ryohei Sasaki, K. Konishi, Tomohiro Takahashi, T. Furukawa","doi":"10.23919/EUSIPCO.2017.8081224","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081224","url":null,"abstract":"This paper proposes a new algorithm for image inpainting algorithm based on the matrix rank minimization with nonlinear mapping function. Assuming that each intensity value of a nonlinear mapped image can be modeled by the autoregressive (AR) model, the image inpainting problem is formulated as a kind of the matrix rank minimization problem, and this paper modifies the iterative partial matrix shrinkage (IPMS) algorithm and provides an inpainting algorithm, which estimates a nonlinear mapping function and the missing pixels simultaneously. Numerical examples show that the proposed algorithm recovers missing pixels efficiently.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129126528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081691
M. Preti, D. Ville
Dynamic functional connectivity (dFC) derived from resting-state functional magnetic resonance imaging (fMRl) allows identifying large-scale functional brain networks based on spontaneous activity and their temporal reconfigurations. Due to limited memory and computational resources, these pairwise measures are typically computed for a set of brain regions from a pre-defined brain atlas, which choice is non-trivial and might influence results. Here, we first leverage the availability of dynamic information and new computational methods to retrieve dFC at the finest voxel level in terms of dominant patterns of fluctuations, and, second, we demonstrate that this new representation is informative to derive meaningful brain parcellations that capture both long-range interactions and fine-scale local organization. We analyzed resting-state fMRI of 54 healthy participants from the Human Connectome Project. For each position of a temporal window, we determined eigenvector centrality of the windowed fMRl data at the voxel level. These were then concatenated across time and subjects and clustered into the most representative dominant patterns (RDPs). Each voxel was then labeled according to a binary code indicating positive or negative contribution to each of the RDPs. We obtained a 36-label parcellation displaying long-range interactions with remarkable hemispherical symmetry. By separating contiguous regions, a finer-scale parcellation of 448 areas was also retrieved, showing consistency with known connectivity of cortical/subcortical structures including thalamus. Our contribution bridges the gap between voxel-based approaches and graph theoretical analysis.
{"title":"Fine-scale patterns driving dynamic functional connectivity provide meaningful brain parcellations","authors":"M. Preti, D. Ville","doi":"10.23919/EUSIPCO.2017.8081691","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081691","url":null,"abstract":"Dynamic functional connectivity (dFC) derived from resting-state functional magnetic resonance imaging (fMRl) allows identifying large-scale functional brain networks based on spontaneous activity and their temporal reconfigurations. Due to limited memory and computational resources, these pairwise measures are typically computed for a set of brain regions from a pre-defined brain atlas, which choice is non-trivial and might influence results. Here, we first leverage the availability of dynamic information and new computational methods to retrieve dFC at the finest voxel level in terms of dominant patterns of fluctuations, and, second, we demonstrate that this new representation is informative to derive meaningful brain parcellations that capture both long-range interactions and fine-scale local organization. We analyzed resting-state fMRI of 54 healthy participants from the Human Connectome Project. For each position of a temporal window, we determined eigenvector centrality of the windowed fMRl data at the voxel level. These were then concatenated across time and subjects and clustered into the most representative dominant patterns (RDPs). Each voxel was then labeled according to a binary code indicating positive or negative contribution to each of the RDPs. We obtained a 36-label parcellation displaying long-range interactions with remarkable hemispherical symmetry. By separating contiguous regions, a finer-scale parcellation of 448 areas was also retrieved, showing consistency with known connectivity of cortical/subcortical structures including thalamus. Our contribution bridges the gap between voxel-based approaches and graph theoretical analysis.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123925721","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081550
M. B. Kocamis, F. Oktem
Near-field ultrawideband imaging is a promising remote sensing technique in various applications such as airport security, surveillance, medical diagnosis, and through-wall imaging. Recently, there has been increasing interest in using sparse multiple-input-multiple-output (MIMO) arrays to reduce hardware complexity and cost. In this paper, based on a Bayesian estimation framework, an optimal design method is presented for two-dimensional MIMO arrays in ultrawideband imaging. The optimality criterion is defined based on the image reconstruction quality obtained with the design, and the optimization is performed over all possible locations of antenna elements using an algorithm called clustered sequential backward selection algorithm. The designs obtained with this approach are compared with that of some commonly used sparse array configurations in terms of image reconstruction quality for various noise levels.
{"title":"Optimal design of sparse MIMO arrays for near-field ultrawideband imaging","authors":"M. B. Kocamis, F. Oktem","doi":"10.23919/EUSIPCO.2017.8081550","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081550","url":null,"abstract":"Near-field ultrawideband imaging is a promising remote sensing technique in various applications such as airport security, surveillance, medical diagnosis, and through-wall imaging. Recently, there has been increasing interest in using sparse multiple-input-multiple-output (MIMO) arrays to reduce hardware complexity and cost. In this paper, based on a Bayesian estimation framework, an optimal design method is presented for two-dimensional MIMO arrays in ultrawideband imaging. The optimality criterion is defined based on the image reconstruction quality obtained with the design, and the optimization is performed over all possible locations of antenna elements using an algorithm called clustered sequential backward selection algorithm. The designs obtained with this approach are compared with that of some commonly used sparse array configurations in terms of image reconstruction quality for various noise levels.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124178128","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-08-01DOI: 10.23919/EUSIPCO.2017.8081669
Qiang Li, Chao-xu Li, Jingran Lin
The multi-input single-output multi-eavesdropper (MISOME) wiretap channel is one of the generic wiretap channels in physical layer security. In Khisti and Wornell's classical work [1], the optimal secure beamformer for MISOME has been derived under the total power constraint. In this work, we revisit the MISOME wiretap channel and focus on the large-scale transmit antenna regime and the constant modulus beamformer design. The former is motivated by the significant spectral efficiency gains provided by massive antennas, and the latter is due to the consideration of cheap hardware implementation of constant modulus beamforming. However, from an optimization point of view, the secrecy beamforming with constant modulus constraints is challenging, more specifically, NP-hard. In light of this, we propose two methods to tackle it, namely the semidefinite relaxation (SDR) method and the ADMM-Dinkelbach method. Simulation results demonstrate that the ADMM-Dinkelbach method outperforms the SDR method, and can attain nearly optimal secrecy performance for the large-scale antenna scenario.
{"title":"Constant modulus beamforming for large-scale MISOME wiretap channel","authors":"Qiang Li, Chao-xu Li, Jingran Lin","doi":"10.23919/EUSIPCO.2017.8081669","DOIUrl":"https://doi.org/10.23919/EUSIPCO.2017.8081669","url":null,"abstract":"The multi-input single-output multi-eavesdropper (MISOME) wiretap channel is one of the generic wiretap channels in physical layer security. In Khisti and Wornell's classical work [1], the optimal secure beamformer for MISOME has been derived under the total power constraint. In this work, we revisit the MISOME wiretap channel and focus on the large-scale transmit antenna regime and the constant modulus beamformer design. The former is motivated by the significant spectral efficiency gains provided by massive antennas, and the latter is due to the consideration of cheap hardware implementation of constant modulus beamforming. However, from an optimization point of view, the secrecy beamforming with constant modulus constraints is challenging, more specifically, NP-hard. In light of this, we propose two methods to tackle it, namely the semidefinite relaxation (SDR) method and the ADMM-Dinkelbach method. Simulation results demonstrate that the ADMM-Dinkelbach method outperforms the SDR method, and can attain nearly optimal secrecy performance for the large-scale antenna scenario.","PeriodicalId":346811,"journal":{"name":"2017 25th European Signal Processing Conference (EUSIPCO)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2017-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124213833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：