Mutluoglu, M.B. | Baydere, S.

Article | 2011 | International Journal of Ad Hoc and Ubiquitous Computing7 ( 4 ) , pp.221 - 234

Wireless Sensor Networks (WSN) monitor large environments for long periods of time. During this period, mechanisms are needed to propagate code updates to the network when needed. In this paper, a distributed code propagation algorithm that utilises reliable transport to all sensors is presented. A mathematical model is developed to analyse the achievable network branch capacities in terms of optimum node numbers per tree level. The outcome of the analysis is used to construct an optimum tree with heuristically best children lists. The capacity analysis is verified with experimental testbed and simulation results and the performance . . . gain is shown. © 2011 Inderscience Enterprises Ltd Daha fazlası Daha az

Uslu, G. | Baydere, S.

Article | 2015 | Expert Systems with Applications42 ( 21 ) , pp.8052 - 8063

Abstract Activity monitoring systems (AMS) detect actions performed by humans. For an AMS to be effectively deployed in daily life, it should partition sensor data streams in real time and determine what activity corresponds to each partition. In this work, a real time continuous activity monitoring system, named RAM, is proposed. RAM detects simple and composite activities, collecting the data with a single 3D accelerometer to produce a non-invasive solution. Classification module of RAM carries out non-predefined feature extraction and activity detection in a coalesced manner thanks to feature extractive training, whereas the stat . . .e-of-art classifiers need to be fed with the output of a predefined feature extraction scheme. As being a Support Vector Machines (SVM) inspired solution, RAM fulfils multiclass classification with one-against-pseudo class strategy, without generating hyperplanes. The strength of the proposed model lies in that RAM achieves robustness in terms of inter-activity detection consistency and time efficiency with little overhead. Robustness property offers a potential to reduce the need for re-training an expert system, which faces the problem of growing set of activity classes in the real time activity recognition domain. We compared RAM for a set of hand oriented activities, against 8 different configurations, where SVM and K-Nearest Neighbour (KNN) classifiers are fed with different predefined features. We observed that RAM outperforms these configurations in overall accuracy as well as inter-activity detection consistency. We also presented the results of real tests as a proof-of-concept for transition detection in composite activities. © 2015 Elsevier Ltd Daha fazlası Daha az

Demir, A.K. | Demiray, H.E. | Baydere, S.

Article | 2014 | Wireless Networks20 ( 4 ) , pp.655 - 670

Wireless multimedia sensor networks (WMSNs), having inherent features and limited resources, require new quality of service (QoS) protocols for real-time and multimedia applications. In this paper, we present a cross-layer QoS architecture (QoSMOS), that unifies network and link layers into a single communication module for QoS provisioning. Based on QoSMOS architecture, we developed an example reference cross-layer protocol, named cross-layer communication protocol (XLCP), enabling scalable service differentiation in WMSNs. Comprehensive analysis of simulation results indicate that the proposed architecture successfully differentia . . .tes service classes in terms of soft delay, reliability and throughput domains. A comparative analysis of XLCP and its counterparts is also given to show the superiority of the cross-layer protocol. © 2013 Springer Science+Business Media New York Daha fazlası Daha az

Turkes, O. | Baydere, S.

Conference Object | 2011 | Proceedings of the 2011 7th International Conference on Intelligent Sensors, Sensor Networks and Information Processing, ISSNIP 2011 , pp.317 - 322

Audio data processing and transmission in Wireless Sensor Networks (WSNs) require systems which satisfy the equilibrium between heavy data traffic and limited resources. In the lossy nature of the network, transmission techniques matter to sustain a certain content validity to be preserved. Besides, data handling is a worth-stressing issue to sustain a decreased computation and traffic overhead. An admissible quality of the voice signals being gathered at the sink node is to be directly maintained with network transmission assessments. To this respect, basic characteristics of a voice signal must be compromised with network properti . . .es. In this study, we investigate the quality of voice signals sent through the homogeneously constructed multi-hop network in which recorded signal segments in the source node are touched with the counterparts expected to be collected in each hop with a transmission rating factor. The basic characteristics of the voice samples used in the testbed are essayed with different in-network qualifications that can directly affect the quality; so that a reasonable trade-off between voice quality and sensor network capabilities is tried to be proposed. © 2011 IEEE Daha fazlası Daha az

Uslu, G. | Baydere, S.

Conference Object | 2017 | 2016 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2016 - Conference Proceedings , pp.3528 - 3533

In real time continuous activity recognition systems, utilization of a data segmentation stage increases the dependency of success ratio on the size of activity set as well as activity type, duration and sensor sampling rate. In this study, we analyzed if iterative K-Nearest Neighbour based knowledge discovery performed on acceleration data can substitute for the segmentation stage to reduce these dependencies for hand oriented activities. To this end, we compared peak frequency and wavelet entropy feature extraction schemes for the recognition of open-pill-box, put-pill-in-mouth, drink and put-glass-back actions which constitute as . . . a whole 'medication intake' activity. We evaluated the performance of these schemes on incomplete data, resulting from the iterative process. According to our findings, peak frequency outperforms wavelet entropy in terms of Intra-Class-Correlation (ICC) metric which is an indication for better adaptation to variation in activity type with incompleteness mitigation problem. Our work points out that a more proficient iterative classification algorithm is required for attaining higher adaptability to the diversification of actions that stem from the incompleteness problem. © 2016 IEEE Daha fazlası Daha az

Güney, I.A. | Ovant, B.S. | Baydere, S.

Conference Object | 2016 | 2016 International Conference on High Performance Computing and Simulation, HPCS 2016 , pp.350 - 356

Remote Direct Memory Access(RDMA) technology offers a promising solution to the memory wall problem in high speed communication. In this paper, we analyze the performance of RDMA and alternative communication models for the utilization of distributed breadth first search (BFS) traversal on large full tree graphs. First, we present the scalability limitations of the shared-memory multi-threading solution along with a sequential benchmark. We then compare the performance of socket-based distributed and RDMA-based distributed communication models on nine different graph structures. The results illustrate that distributed solutions are . . .more viable than shared memory models for the given test cases. Among the distributed models, RDMA achieves 28% lower latency than the socket communication for graphs having 22.3M vertices on eight peer nodes. RDMA also achieves 98% lower latency than the sequential algorithm. © 2016 IEEE Daha fazlası Daha az

Sarisaray-Boluk, P. | Gungor, V.C. | Baydere, S. | Harmanci, A.E.

Article | 2011 | Ad Hoc Networks9 ( 7 ) , pp.1287 - 1301

With the recent advances in Micro Electro-Mechanical Systems (MEMS) and underwater imaging sensors, and cameras, underwater multimedia sensor networks (UMSNs) have been proposed and drawn the immediate attention of the research community. Underwater multimedia sensor networks enable several new applications, such as target tracking, advanced coastal multimedia surveillance, undersea explorations, image acquisition and classification, environmental monitoring, and disaster prevention. However, the practical realization of these currently designed and envisioned applications directly depends on reliable and quality-aware communication . . . capabilities of the deployed UMSNs. This paper presents a comprehensive performance evaluation of error concealment and error correction algorithms for quality-aware image transmission over UMSNs. Specifically, different combinations of multipath transport,watermarking-based error concealment (EC), forward error correction (FEC) and adaptive retransmission mechanisms have been evaluated to combat underwater channel impairments and mitigate packet losses due to node failures and intrinsic underwater acoustic channel characteristics. In addition, two novel image quality assessment metrics have been proposed to obtain the predicted quality of the image depending on the channel and node failures. Comparative performance evaluations show that the EC approach reconstructs the distorted image as closely as the original one while avoiding the burden of retransmissions and consequent delay. © 2011 Elsevier B.V. All rights reserved Daha fazlası Daha az

Uslu, G. | Baydere, S.

Conference Object | 2013 | 2013 21st Signal Processing and Communications Applications Conference, SIU 2013 , pp.1287 - 1301

Human activity monitoring enables detecting instances when people need help during daily routines. They may have forgotten taking medication or they can experience more severe situations such as falling. Detecting their activities yield their context information revealing occurrences of such cases. We designed and implemented a solution to activity detection proposing a Support Vector Machine (SVM) based method. We gathered data through accelerometer to come up with a noninvasive solution. Our method is the combination of a feature extractor and classifier. Presented activity recognition suit eliminates the need for experimenting wi . . .th multiple features to determine the best classifying features contrary to some approaches utilizing SVM. With our SVM based activity recognizer, we classified sit, stand, lie and walk actions with 100 % accuracy. © 2013 IEEE Daha fazlası Daha az

Baydere, S. | Çayirci, E. | Hacioglu, I. | Ergin, O. | Ollero, A. | Maza, I. | Lijding, M.

Book Part | 2010 | Cooperating Embedded Systems and Wireless Sensor Networks , pp.25 - 113

[No abstract available]

Turkes, O. | Baydere, S.

Conference Object | 2012 | Proceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM , pp.35 - 42

Recently, multimedia utilization inWireless Sensor Networks (WSN) has shown that robust encoding methods are imperative for any application requiring a certain level of quality. During ubiquitous data exploitation in these lossy networks, a data-conserving method coherent with coding and transmission scheme is essential. This study centers upon several basic reconstruction methods for unapprehended parts in the voice data gathered at the end-point of a real multi-hop Audio Sensor Network (ASN). Considering gathered voice signals, error concealment (EC) methods are inherently applied over lost packets in the testbed. Around 6,000 rea . . .l single-path transmission tests are verified with an instrumental simulation. Besides, EC schemes are supported with a multi-path transmission in which data aggregation occurs at certain intervals. Nearly 300,000 qualitative results show that perceptual quality can be preserved promisingly with the utilization of low cost affordable correction techniques. Copyright © 2012 ACM Daha fazlası Daha az

Can, O. | Küçük, G. | Baydere, S. | Sarpkaya, A.O. | Gürsoy, O.

Conference Object | 2015 | 2015 23rd Signal Processing and Communications Applications Conference, SIU 2015 - Proceedings , pp.723 - 726

Today, availability of high-speed networks and improved storage and processing power on servers enable Video on Demand (VoD) services much more accessible than before. The VoD services offer very high flexibility to clients and let them watch any video content whenever they like. In this study, we present the design of a scalable VoD server that is capable of supporting in the order of thousands video requests on a local area network. Our proposed design does not require a set-top box at the client side. After its design and implementation of the system, a series of tests are carried out with both 360p and 720p video streams, and CP . . .U, memory and I/O statistics are analyzed. The results show that the major bottleneck in the VoD server is the network bandwidth. When it is provided to be as large as 1 GB/s, we show that our proposed VoD server is capable of serving thousand clients. © 2015 IEEE Daha fazlası Daha az

Kosucu, B. | Irgan, K. | Kucuk, G. | Baydere, S.

Conference Object | 2009 | Proceedings of the 2009 ACM International Wireless Communications and Mobile Computing Conference, IWCMC 2009 , pp.1173 - 1177

Wireless sensor networks promise great success in many areas from environmental monitoring to medical and military applications. Forest fire detection is one of these areas where many of the ongoing WSN research is focused today. Unfortunately, most of these studies choose simulating their proposed solutions instead of doing experiments in real testbed environments, since that kind of setup exposes additional difficulties. Our previous work, named FireSense, proposed a fire detection algorithm, which was shown to be successful in terms of simulation results. In this study, we take FireSense to a real outdoor testbed for further anal . . .ysis of its effectiveness in terms of various parameters such as link and node failures, topology and physical configuration changes, wind direction, ignition point position and sampling period variations. Copyright © 2009 ACM Daha fazlası Daha az