Read e-book Optimal Mobile Sensing and Actuation Policies in Cyber-physical Systems

Free download. Book file PDF easily for everyone and every device. You can download and read online Optimal Mobile Sensing and Actuation Policies in Cyber-physical Systems file PDF Book only if you are registered here. And also you can download or read online all Book PDF file that related with Optimal Mobile Sensing and Actuation Policies in Cyber-physical Systems book. Happy reading Optimal Mobile Sensing and Actuation Policies in Cyber-physical Systems Bookeveryone. Download file Free Book PDF Optimal Mobile Sensing and Actuation Policies in Cyber-physical Systems at Complete PDF Library. This Book have some digital formats such us :paperbook, ebook, kindle, epub, fb2 and another formats. Here is The CompletePDF Book Library. It's free to register here to get Book file PDF Optimal Mobile Sensing and Actuation Policies in Cyber-physical Systems Pocket Guide.

Actuation guidelines are then brought into the framework with the aim of bettering estimation and optimizing the trajectories of either sensors and actuators at the same time. The monograph concentrates at the use of equipment for which a cyber-physical-systems infrastructure is needed. The equipment are computationally heavy and require cellular sensors and actuators with communications talents. Download e-book for iPad: Home Networking Annoyances by Kathy Ivens.

The great thing about a house community is that it could make existence really easy - what will be greater than sharing a web connection in order that every person will be on-line whilst? Optimal Observation for Cyber-physical Systems addresses the challenge, fundamental to the design of WSNs, presented by the obligatory trade-off between precise estimates and system constraints. A unified theoretical framework, based on the well-established theory of optimal experimental design and providing consistent solutions to problems hitherto requiring a variety of approaches, is put forward to solve a large class of optimal observation problems.

The Fisher information matrix plays a key role in this framework and makes it feasible to provide analytical solutions to some complex and important questions which could not be answered in the past. Finally, the correlation-based mechanism exploits relationships in spatial and temporal domains between different nodes to reduce redundant data transmissions.

Data dissemination in the sensor network domain of CPSN is an important aspect. To enable consistent communication reliability in CPSN sensing plane throughout the monitoring period, a query-and-reply method between nodes is used frequently that may become difficult to handle in complex topologies with thousands of nodes. This step of data dissemination with communication reliability can thus be carried out in more intelligent ways all of which can be broadly categorized under query-and-response.

  • Blood Type B Food, Beverage and Supplement Lists?
  • "Optimal Sensing and Actuation Policies for Networked Mobile Agents in " by Christophe Tricaud.
  • Mikrocontroller und Mikroprozessoren ( (German Edition).
  • Controversies in Local Economic Development: Stories, Strategies, Solutions (Regions and Cities)?
  • The Differentiated Instruction Book of Lists (J-B Ed: Reach and Teach).

We mainly classify query-and-response based data dissemination into three categories: Directed diffusion is a data centric information dissemination method. Data generated by sensor nodes is given in the form of attribute-value pairs. A node requests data by sending interest messages for named data.

Data matching the interest is then drawn and routed towards that node. Intermediate nodes that fall in the data path can cache, or transform data, and may even direct interests based on previously cached data information. Distributed indexing on the other hand uses a distributed addressing method to gather information. For more complex topologies, hybrid query resolution approaches may be adopted that are a mixture of diffusion and address indexing methods.

For Authors

Avoiding disconnections and providing extensive coverage is another essential design parameter that affects the system performance. In a simple CPSN sensing node with plane layouts and theoretical design, sensors may be assumed to have fixed communication and sensing ranges initially. In a simplistic manner, the coverage problem for CPSN can be formulated as a decision problem where, given a number of sensors to be deployed in the field, the main target is to determine if the area is sufficiently n -covered, i.

Sensing plane coverage solutions for CPSN can be categorized into opportunistic sensor node selection mechanisms or covering-set methods that utilize graph theory. Finally, sensor node localization can be done using Global Positioning System GPS related methods for outdoor deployments or through trilateration, proximity and other out-of-range methods including fixed anchor nodes Figure 4. One important network parameter that concerns CPSN deployment is the node mobility that needs to be properly accounted for in the networking information for dynamic link perspectives and mobile robots deployed over the monitoring area.

Mobility allows network capability to be improved in many ways, e. Overall, such mobility related solutions for CPSN can be classified into two types where the first one would be to try relocating sensor nodes to improve network coverage and connectivity while the other solution would be to try addressing the path planning issues for data relaying nodes that would ultimately allow extended network lifetime [ 16 ].

Highlights of the mobility and reconfiguration related parameters, issues and research dimensions for CPSN are listed in Figure 5. Wireless sensor network parameters and issues related to mobility and reconfiguration in a cyber-physical environment. QoS provisioning becomes a much complex task when the sensing plane of CPSN needs to address thousands of sensor nodes placed in various topologies [ 17 ].

These challenges need to be addressed on both sides of the gateway, i. SOA plays an important role in reducing the complexity of the infrastructure by decomposing CPS functions into smaller distinguishable units each viewed as a separate service [ 18 ]. This allows rapid, efficient and scalable development of a CPS application through reusable service units. Application level QoS, however, needs to be defined according to the service for intelligent cross layered communication.

With the future trends in user applications [ 19 ], CPSN environments would require dynamic system settings for unpredictable environments [ 20 ]. Self-management policies would be needed so that allocated resources like CPU time, bandwidth memory, energy profiles can be controlled intelligently in a high level QoS constrained system setup [ 21 , 22 ].

A major concern with QoS provisioning also pertains to the minimization of energy consumption for major network elements.

1. Introduction

In this regard, Cloud Computing provides a possible solution wherein major enabling technologies for such a setup are Virtualization and Ubiquitous connectivity. Virtualization related technologies like Software Defined Networking SDN and virtual operating systems and connections allow provision of dynamically changing and altering resources based upon service isolation, thus enabling scaling and managing of resources in a more controlled way. Since the resources of the cloud connecting the sensing platform are dynamically manipulated, the cloud itself would provision the typical three types of services to the sensor platform on one side and the remote user on the other side.

SaaS provides services to remote users on a demand basis. PaaS provides a development environment that is encapsulated and offered to users as a service wherein higher level applications can work over it. Finally, IaaS is responsible for provisioning computing capabilities and basic storage as standardized services for both sides of the network. Once the sensing platforms run under a cloud, several differentiations can be made according to the sensing application in hand.

Since integrating WSNs with the cloud makes it easy to share and analyze real time sensor, further advantage can be taken by provisioning sensor data or sensor event as a service over the internet; hence, the terms Sensing as a Service SaaS and Sensor Event as a Service SEaaS are sometimes used.

Example implementation layers and protocols for a cyber-physical sensor network environment. Prediction of accurate output decisions and reliability of sensing information are considered critical for CPSN systems. There is therefore a need to define strictly the network requirement factors in terms of cyber and physical domains Figure 6. These factors also form the QoS basis for achieving a real time intelligent system for high stress and constrained environments like mining, healthcare and warfare.

QoS factors like seamless data flows through the cloud and timely delivery at the monitoring station are considered critical for cyber systems. This becomes more challenging when CPS is integrated with other technologies like semantic agents and hybrid system states in the Cloud. Deployment of CPSN architectural parts require placement of sensing and actuator devices at strategically critical points with intelligent algorithms for node localization and geo-location detection.

The Medium Access Control MAC at the sensing side should consider that the negotiation between neighboring data collection devices and sensors must conserve resources like bandwidth, number of channels, buffer storage and transmission energy [ 23 ]. Software, hardware, middleware and operating systems are required that provide reliability beyond existing technologies. The software and hardware must be highly dependable, certifiable, configurable and, where required, be able to fully integrate with complex systems. A complex CPSN system must possess reliability that is currently lacking in many deployments.

Though overdesign is currently the safest path for system deployment, this approach becomes intractable for complex designs and systems where interoperability is required. Evidence based methods are needed for reasoning inference about system reliability. New methods, algorithms, models and tools are needed that can incorporate verification and validation of software and systems at the control stage.

A major focus must be on systems which are verifiably robust in order to keep operating under situations and environments with uncertainty combined with potentially rapidly changing objectives. CPSNs built for optimization, control and scheduling will have to interoperate efficiently in real time. Hence, further study is required to explore the organizing principles of such interactions and appropriate abstractions that can support services with significantly shortened design cycles.

Before CPSN implementation, it is important to develop representative models of all agents including electrical, mechanical and computational components as well as important environmental factors which influence the dynamics of the system. With the availability of a high level of computational power, several tools have also emerged for direct modeling of dynamic systems. Issues that dominate the reliability and prediction design requirements include: For monitoring remote sensor applications over the IP framework, cloud computing can provide a middleware cost effective solution to CPSN that provides a rich interactive communication platform.

Since network communication costs a lot of bandwidth overhead for linking Virtual Machines VMs in data intensive environments, a decentralized approach, where migration of VM services is provided with monitoring of traffic fingerprints can relieve the wasted overhead.

Also, in particular cases, faults can occur in the middle of a query from distributed databases. This can be fixed by dividing queries into sub-queries and mapping them in an intelligent way such that the results return on different nodes. A global middleware concept can be a convenient way to provide flexibility integration and discovery of sensor networks and sensor data. Such a middleware would be required to provide fast deployment of testbeds with distributed querying, filtering and aggregation of sensor data with support for dynamic adaptation of the system configuration parameters.

There was a problem providing the content you requested

This would be linked to the use of virtual sensor abstraction that can enable users to declaratively specify deployment descriptions in an open standard human readable language like XML. Such an approach becomes more powerful for remote monitoring once there is a possibility of integration of sensor network data through querying language like SQL over local and remotely available sensor network resources. In addition to SML, XML encoding will be used for the measurement and description of the physical sensor specifications. The use of XML language allows encoding the sensors in a way that the implementation is available across several hardware platforms and operating systems through simple translation or use of a wrapper.

A map for translating between physical and virtual sensor parameters can be used to translate commands. The concept of metadata in Sensor-Cloud perspective holds the importance of publishing the type and location of the sensor at the time of data generation. In CPSN, location for different data generating and terminating points would serve as a first class knowledge for many relevant applications [ 24 ].

Fler böcker av författarna

Compared to outdoor location detection through GPS and similar approaches, indoor location estimation or localization proves more challenging. Typical systems proposed in this context such as the Bat and Cricket [ 25 ] relate to smaller scale environments, while to cope with newer demands of extended scaled up systems, pattern matching of data can be applied as a useful solution.

Locating the event on the relevant sensor node when requested by a remote application is an important issue. Several methods have been proposed in the literature to account this issue. One method is to locate the physical sensors with data faults by assuming a mismatch between the sensor data rank and the distance rank. The command translating agent at the gateway side linked to the Cloud would directly depend upon the CPSN programming scope for implementation of remote sensor network monitoring using cyber physical systems [ 26 ].

Major aspects in the middleware programming scope are highlighted in Figure 7 where the code implementation can be done in a number of ways depending upon the sensor data and processing requirements. Typical sensor data processing implementations include sequential, event driven, functional and rule based filtering. An example of typical programming instance to be run on data sensing model is given in Figure 8 that uses an event driven approach to detect a sensor value that is categorized as a leak.

The data access model for connecting the data base can be implemented using several combinations of steps like message parsing and use of mobile code for timely data retrieval Figure 7. CPSN is actually a bridge to link the cyber world with communication, intelligence and information components with the physical world counterpart providing sensing and actuation capabilities [ 27 ]. The CPSN platform may be broadly classified as an integration of an intelligent control design system with a mobile or static sensor or actuator system.

When considering standalone individual sensor networks, issues like network formation, security, mobility and power management remain almost the same in a broader perspective. However, major technical differences for the CPSN approach include the use of heterogeneous information flow, multi-dimensional sensor cooperation and a high level of intelligence and algorithms informing the actuation and decision framework. For example, a complete CPSN can be used to assist in management of greenhouse sensing information at an extremely large geographical distance.

More complex systems could include multiple sensors and actuators that can be used for applications such as environment-related climate control settings with humidity, heating, carbon dioxide generation, fertilizing and watering system features [ 28 ] Table 4. Summary of wireless sensor network based cyber physical monitoring of pipeline infrastructure.

Important elements for CPSN based pipeline infrastructure monitoring require intelligent sensor integration according to the pipeline layout, correct sensor event detection algorithms, data fusion and inference techniques, data routing across the network and cloud interfaces. Pipelines can be underground as well as above ground; hence, sensors calibration and positioning would be affected by atmospheric phenomenon like winds, humidity and day period. Quite recently, the use of acoustic sensors for underwater environment monitoring has become popular where several research groups are actively deploying them in experimental testbeds like coral reef monitoring and underwater pipelines [ 29 , 30 , 31 ].

Once the sensors are in place, the sensing algorithms play an important role in timely detection of sensor events using reactive, proactive or mixed algorithm trigger mechanisms. Once the events are detected from multiple sensors, diverse information on a node can be fused at sensor node for compressing data following the sensor network data rates being typically low. The data can also be fused at relay or gateway nodes in case the CPSN encompasses several sub sensor networks. In addition to data fusion, data processing codes can be found in a distributed manner where one code instance runs on the sensor nodes and another runs on the remote control station.

Overall, design of a CPS based wireless sensor network would require topology controlled infrastructure design, actuation mechanisms, an intelligent middleware lying in the Cloud and a data routing mechanism from the sensor node towards the remote monitoring station, as well as feedback from the controller to the actuation platform. Oil and gas pipeline monitoring provides a novel example where WSN can provide remote monitoring while CPS integration can be leveraged to apply real time information and analytics of the underlying framework [ 32 ].

With a number of possible pipeline deployment techniques available, all efforts unswervingly reflect the characteristics of the medium needed for transportation that depends upon the environmental, strategic and economic conditions. For intelligently providing flexible and reliable CPSN-based monitoring, factors like sensor layout, data transmission methods, sensor node power concerns, data processing, analysis and inference points, operational design and framework in addition to network topology, infrastructure and sensing related technologies are the focus of this section.

The oil and gas distribution pipeline deployment methods can be widely characterized into underwater, above ground and buried pipelines. Since leakages could be detrimental to the surroundings, integral pipeline monitoring is essential and needs to be reliable and in real time. The main sensing techniques can be categorized into reactive ones that detect the presence of leaks only once they occur while the proactive methods monitor the condition of pipelines gradually over time to prevent leakage occurrence [ 33 ].

The type of sensors, their placement and their usage in the monitoring environment form the basis for CPSN monitoring systems. Sensors placement for pipeline monitoring can be classified as outside placement and inside placement that may use invasive or pervasive techniques depending on the scenarios given [ 34 ]. Monitoring changes in pipelines with visual inspection is a handy technique for monitoring events that occur above ground.

Vision sensors allow distinguishing differences or changes in the area around the pipeline. With such sensors, small changes in the physical nature of the pipelines, temperature difference or the event of any fluid or gas leakage can be easily determined. Ground-penetrating radar technique is another widely used method to accurately monitor changes and collect evidence of the existence of any occurrences at the ground level without digging.

With the help of acoustic transducers, small fluid or gas leaks are easily identified as they produce frequency oscillations [ 35 , 36 ]. The coverage range of these transducers is very small; hence, a number of transducers are required to cover the desired underground pipeline requiring extensive operational and maintenance efforts. To reduce such efforts, acoustic transducers are only deployed on the pumping stations or near checkpoints. With the mass balancing method [ 36 ], the difference in flow of entering and leaving fluid can be monitored though it is not an efficient method for locating major leakages in underground pipelines.

One of the ways to detect leaks is by comparing the change in temperature profiles of the immediate surroundings of the pipeline due to the Joule Thompson Effect [ 37 ]. Temperature sensors for this purpose should be selected keeping the temperature gradient in mind. Distributed fiber optic sensors [ 38 ] provide a temperature signature and pinpoint the leak location. Optical fiber sensors need to be in close contact with the pipe so as to come in contact with a leaked fluid Figure Figure 10 illustrates an optical fiber-based leakage detection mechanism for gas pipelines using a distributed acoustic sensing technique that measure the backscatter that results due to disturbance caused in the fiber.

The normal temperature profile and the temperature profile sensed from the fiber sensor in the presence of a leak are also illustrated. The fiber sensor is placed above the pipeline for gases and below the pipeline for liquids [ 13 ].

Download e-book for kindle: Optimal Mobile Sensing and Actuation Policies in by Christophe Tricaud

This sensor provides an effective solution for pipeline monitoring because of its low cost and easy installation. One attractive feature of optical fiber sensing is the fact that the optical fibers have already been installed and used for communication purposes along the pipelines [ 38 ]. Several techniques have been proposed for under-ground sensing of pipelines. Sensing soil properties can be useful to find abnormalities in soil [ 14 ]. Sensors can detect temperature variation in the soil due to leakage of hot liquids [ 36 , 39 ].

Hydrocarbon vapor sensors can be used to detect the leakage of pipelines transporting liquefied natural gas [ 39 ]. Soil dielectric property sensors can be used to detect the leakage of crude oil pipelines [ 40 , 41 ]. Acoustic sensors can be used to detect leaks in municipal water pipes, sewages, and oil and gas pipelines, respectively [ 29 , 30 , 31 ]. Since corrosion in pipelines makes the inner pipe surface vulnerable to leakages and any external or internal strike like solid particles hitting the surface with high velocities can result in leakages, it is also necessary to monitor pipes flowing fluid for the presence of solid particles.

Using acoustic sensors to detect the presence of solid impurities such as sand in oil, the sensors are placed non-intrusively near the bends of the pipeline [ 42 ]. The solid particles collide with the surface of the pipeline at the bends which generate high frequency waves that are detected by the sensors. Mobile sensor robots move inside the pipeline to monitor pipeline conditions and to provide accurate measurements and readings about the defects appearing inside the pipeline at regular intervals.

Industries are therefore interested in developing robots with ingenious designs so the whole pipeline can be scrutinized [ 43 ]. A recently developed Magnetostrictive Sensor MsS inspection system [ 44 ] is an intrusive sensor which detects corrosion in the pipelines. The research contribution in [ 47 ] identifies construction equipment to be one of the major causes of breakages of pipelines and proposes efficient acoustic sensing with noise cancellation that can detect the presence of such equipment and generate an alarm at the base station.

PipeSense [ 48 ] has provided an alternate sensing method to acoustic sensing by incorporating induction based ad hoc RFID wireless sensor networks for water pipelines. Meanwhile, work described in [ 48 , 49 ] uses pressure signals to measure the events like leakages and bursts by monitoring the signal states using stochastic HMM processes. SWATS [ 50 ], a system for monitoring steam flood and water flood pipelines, makes use of all common measurements such as pressure, flow and temperature.

Work in [ 51 ] describes a thermal video technology for leak detection. Thermal cameras are used to exploit temperature differentials which provide greater accuracy than methods that rely on color and size characteristics. For problems created inside the pipeline, several in-pipeline monitoring applications have been stated in the literature. Such problems resulting in leakages may be caused by sudden changes in pressure, corrosion, cracks, bad workmanship, defects in pipes or lack of maintenance [ 52 ].

Since the detection system comes closer to the location where the leakage is inside the pipeline, the in-pipeline leakage detection methods are considered more accurate and less sensitive to external noise. The Smartball [ 53 ] as a mobile sensor device can be used to detect and locate small leaks inside pipelines that are larger than 6 inches in diameter. The sensor device is developed in the form of a free swimming device consisting of a porous foam ball that envelopes an aluminium sphere containing the sensitive acoustic instrumentation. Another in-pipeline monitoring device, Sahara [ 54 ], is able to estimate and locate the leak in large diameter water pipelines.

The system travels by the flow of the water and, in case of a leak, the exact location is marked on the surface by an operator that follows the machine movement. Both Sahara and Smartball are passive since they cannot be actuated and cannot actually maneuver inside the pipeline.

ViewInside - Optimal Mobile Sensing and Actuation Policies in Cyber-physical Systems

Several crawlers have also been reported that utilize wheeled platforms, cameras, and a mechanism for control and communication, e. For in-pipeline leak detection involving gas, the Explorer [ 56 ] has been used which is a long range leak inspection robot controlled by a human operator via wireless RF signals. The explorer constantly looks into an installed camera to search for leaks that are useful for offline inspection. For oil pipelines, nondestructive inspection is the most successful technique. Magnetic flux leakage based detectors and ultrasounds are used quite commonly to search for pipe defects.

Major fault with such an approach is the dependency on pipe material and high power usage. These are also not suitable for long range mission where maneuvering capabilities are difficult due to large pipeline size. PipeGuard is able to detect leaks in a more reliable and autonomous fashion as compared to passive approaches [ 52 ].

The PipeGuard system is inserted into the network through special insertion points and the system reports leakages wirelessly through relay stations.