عنوان مقاله
مدلسازی شبکه های سنسور بی سیم مقاوم در برابر خطا به روش مارکو
فهرست مطالب
چکیده
مقدمه و انگیزه
کارهای وابسته
مدلهای مقاوم در برابر خطا مارکو
نتایج
نتیجه گیری
بخشی از مقاله
کارهای وابسته
اگرچهFT عمومی رشته پژوهش مطالعه شده است اما کارهای کمی در زمینه خطا یابی خاصWSN وFT انجام شده است. Jiang طرح DFD را پیشنهاد نمود که گره های سنسور معیوب را از طریق مبادله داده ها و تست متقابل در میان گره های همسایه، کشف نمود.Jian-Liang طرح (WMFDS) را پیشنهاد کردند که از همبستگی های فضایی در میان اندازه گیریهای سنسور استفاده می کرد (مثلاً دما، رطوبت).
کلمات کلیدی:
Markov Modeling of Fault-Tolerant Wireless Sensor Networks Arslan Munir and Ann Gordon-Ross Department of Electrical and Computer Engineering University of Florida, Gainesville, Florida, USA e-mail: {amunir@ufl.edu, ann@ece.ufl.edu} Abstract—Technological advancements in communications and embedded systems have led to the proliferation of wireless sensor networks (WSNs) in a wide variety of application domains. One commonality across all WSN application domains is the need to meet application requirements (e.g., lifetime, reliability, etc.). Many application domains require that sensor nodes be deployed in harsh environments (e.g., ocean floor, active volcanoes), making these sensor nodes more prone to failures. Unfortunately, sensor node failures can be catastrophic for critical or safety related systems. To improve reliability in such systems, we propose a fault-tolerant sensor node model for applications with high reliability requirements. We develop Markov models for characterizing WSN reliability and MTTF (Mean Time to Failure) to facilitate WSN application-specific design. Results show that our proposed fault-tolerant model can result in as high as a 100% MTTF increase and approximately a 350% improvement in reliability over a non-fault-tolerant WSN. Results also highlight the significance of a robust fault detection algorithm to leverage the benefits of fault-tolerant WSNs. Index Terms—Fault-Tolerance, reliability, Markov modeling, wireless sensor networks I. INTRODUCTION AND MOTIVATION Wireless sensor networks (WSNs) consist of spatially distributed autonomous sensor nodes that collaborate with each other to perform an application task. WSN sensor nodes are typically mass produced and are often deployed in unattended and hostile environments making them more susceptible to failures than other systems [1]. Additionally, manual inspection of faulty sensor nodes after deployment is typically impractical. Nevertheless, many WSN applications are mission-critical, requiring continuous operation. Thus, in order to meet application requirements reliably, WSNs require fault detection and fault-tolerance (FT) mechanisms. Fault detection encompasses distributed fault detection (DFD) algorithms which identify faulty sensor readings that indicate faulty sensors. DFD algorithms typically use existing network traffic to identify sensor failures and therefore do not incur any additional transmission cost. A fault detection algorithm's accuracy signifies the algorithm's ability to accurately identify faults. Though fault detection helps in isolating faulty sensors, WSNs require FT to reliably accomplish application tasks. One of the most prominent FT techniques is to add hardware and/or software redundancy to the system [2]. However, WSNs are different from other systems as they have stringent constraints and the added redundancy for FT must justify the additional cost. Studies indicate that sensors (e.g., temperature and humidity sensors) in a sensor node have comparatively higher fault rates than other components (e.g., processors, transceivers) [3][4]. Fortunately, sensors are cheap and adding spare sensors contribute little to the individual sensor node's cost. Even though FT is a well studied research field [5][6][7][8], fault detection and FT for WSNs are relatively unstudied. Additionally, fault detection and FT for WSNs have added complexities due to varying FT requirements across different applications. For instance, mission critical applications (e.g., security and defense systems) have very high reliability requirements whereas non-mission critical applications (e.g., ambient conditions monitoring applications) typically have relatively low reliability requirements. To the best of our knowledge there exists no sensor node model to provide better reliability for such critical applications. Furthermore, applications are designed to operate reliably for a certain period of time (i.e., WSN applications typically have specific lifetime requirements). Unfortunately, literature provides no rigorous mathematical model with insights into WSN reliability and lifetime. Finally, fault detection and FT have been studied in isolation and their synergistic relationship has not been investigated in the context of WSNs. Our main contributions in this paper are: • We investigate the synergy of fault detection and FT for WSNs and propose an FT sensor node model consisting of duplex sensors (i.e., one active sensor and one inactive spare sensor), which exploits this synergy between fault detection and FT. Whereas sensors may employ Nmodular redundancy (e.g., triple modular redundancy (TMR) is a special case of N-modular redundancy) [2], we propose a duplex sensor model to minimize the additional cost for our FT model. • To the best of our knowledge, we for the first time develop a Markov model for characterizing WSN reliability and MTTF. Our Markov modeling facilitates WSN design by enabling WSN designers to determine the exact number of sensor nodes required to meet the application's lifetime and reliability requirements. Our Markov modeling provides an insight on the type of sensor nodes (duplex or simplex) feasible for an application to meet the application's requirements.
