Real-Time Systems. Design Principles for Distributed Embedded Applications. Herman Kopetz. Second Edition (811374), страница 90
Текст из файла (страница 90)
The important topic of safety and security in the IoT is furtheraddressed in the final section of this chapter.13.3.2 Naming and IdentificationThe vision of the IoT (that all of the billions of smart objects can communicate viathe Internet) requires a well-thought-out naming architecture in order to be able toidentify a smart object and to establish an access path to the object.Every name requires a named context where the name can be resolved. Therecursive specification of naming context leads to a hierarchical name structure –the naming convention adhered to in the Internet.
If we want a name to beuniversally interpretable without reference to a specific naming context, we needa single context with a universally accepted name space. This is the approach takenby the RFID community, which intends to assign an Electronic Product Code(EPC) to every physical smart object (see also Sect. 13.4.2). This is more ambitiousthan the forerunner, the optical bar code, which assigns a unique identifier only to aclass of objects.Isolated Objects.
The following three different object names have to be distinguished when we refer to the simple case of an isolated object:lllUnique object identifier (UID) refers to the physical identity of a specific object.The Electronic Product Code (EPC) of the RFID community is such a UID.Object type name refers to a class of objects that ideally have the same properties.
It is the name that is encoded in the well-established optical bar code.Object role name. In a given use context, an object plays a specific role that isdenoted by the object role name. At different times, the same object can play13.3 Technical Issues of the IoT313different roles. An object can play a number or roles and a role can be played bya number of objects.Example: The assumption that all objects that have the same object type name areidentical does not always hold.
Consider the case of an unapproved spare part that hasthe same visible properties and is of the same type as an approved spare part, but is acheaper copy of the approved part.Example: An office key is an object role name for a physical object type that unlocks thedoor of an office. Any instance of the object type is an office key. When the lock in the officedoor is changed, a different object type assumes the role of the office key. A particularoffice key can also unlock the laboratory. It then plays two roles, the role of an office keyand the role of a laboratory key. A master key can open any office – there are thus twodifferent keys that play the same role.Composite Objects. Whenever a number of objects are integrated to form a composite object, a new whole, i.e., new object is created that has an emerging identitythat goes beyond the identities of the constituent objects.
The composite objectresembles a new concept (see Sect. 2.2.1) that requires a new name.Example: George Washington’s axe is the subject of a story of unknown origin in whichthe famous artifact is still George Washington’s axe (a composite object) despite havinghad both its head replaced twice and its handle replaced three times [Wik10].A composite object that provides an emergent service requires its own UID that ishardly related to the UIDs of its constituent parts. The different names, UID, objecttype name, and object role name must be introduced at the level of compositeobjects as well.
Since a composite object can be an atomic unit at the next level ofintegration, the name space must be built up recursively.Which one of the object names, introduced in the above paragraphs, should bethe access points for the communication with the Internet? It will be difficult tomanage the communication complexity if all objects that are contained in multilevelcomposite objects can be accessed anytime at anyplace.It is evident that the introduction of a flat name space for all smart objects of theuniverse, as stipulated by the EPC is only a starting point.
More research on theproper design of name-space architectures in the IoT is needed.13.3.3 Near Field CommunicationThe IoT requires, in addition to the established WLANs (Wireless Local AreaNetworks), short-range energy-efficient WPANs (Wireless Personal Area Networks) in order to enable the energy-efficient wireless access to smart objectsover a small distance. The IEEE 802.15 standard working group develops standardsfor WPAN networks.
Among the networks that are conforming to the 802.15standards are the Bluetooth network and the ZigBEE network.Originally, Bluetooth has been introduced as a wireless alternative to the RS232wire-bound communication channel [Bar07]. Bluetooth, standardized in IEEE31413 Internet of Things802.15.1, defines a complete WPAN architecture, including a security layer. At thephysical level, it achieves a data rate of up to 3 Mbit/s over a distance of 1 m (Class3 – maximum transmission power of 1 mW) to 100 m (Class 1 – maximumtransmission power 100 mW) using the transmission technology of frequencyhopping. Bluetooth allows multiple devices to communicate over a single adapter.The ZigBee alliance is a group of companies that develops a secure WPAN thatis intended to be simpler, more energy efficient, and less expensive than Bluetooth[Bar07]. ZigBee uses high-level communication protocols based on the IEEE802.15.4 standard for low-power digital radios.
ZigBee devices are requested tohave a battery live of more than a year.The NFC (Near Field Communication) standard [Fin03], an extension of theISO/IEC 14443 proximity-card standard, is a short-range high frequency wirelesscommunication technology which enables the exchange of data between devicesover a distance of <20 cm. The technology is compatible with both existingsmartcards and readers, as well as with other NFC devices, and is thereby compatible with the existing contactless infrastructure already in use for public transportation and payment.
NFC is primarily aimed for use in mobile phones.13.3.4 IoT Device Capabilities versus Cloud ComputingSmart objects that have access to the Internet can take advantage of services that areoffered by the cloud (large data centers that provide their services through theInternet).
The division of work between a smart object and the cloud will bedetermined, to a considerable degree, by privacy and energy considerations[Kum10]. If the energy required to execute a task locally is larger than the energyrequired to send the task parameters to a server in the cloud, then the task is acandidate for remote processing. However, there are other aspects that influence thedecision about work distribution: autonomy of the smart object, response time,reliability, and security.13.3.5 Autonomic ComponentsThe large number of smart objects that are expected to populate our environmentrequires an autonomic system management without the need of frequent humaninteractions.
This autonomic management must cover network service discovery,system configuration and optimization, diagnosis of failures and recovery afterfailures, and system adaptation and evolution. There is a need for a multi-levelautonomic management, starting with the fine-grained management of componentsup to the coarse grained management of massive assemblies of components orlarge systems.13.4 RFID TechnologyFig. 13.1 Model of anautonomic component(Adapted from [Hue08])315autonomic componentautonomy managermonitoranalyzesensorsknowledge baseplanexecuteeffectorsmanaged componentFigure 13.1 shows the generic MAPE-K (Monitoring, Analyzing, Planning,Execution with a Knowledgebase) architecture of an autonomic component[Hue08].
An autonomic component consists of two independent fault-containmentunits (FCU), a managed component and an autonomy manager. The managedcomponent can be a single component, a cluster of components, or a part of a largersystem. The autonomy manager consists of a monitor that observes and analyzesinformation about the behavior of the managed component, a planning module thatdevelops and evaluates alternative plans to achieve the stated goals, and finally aninterface to the managed object that allows the autonomy manager to influence thebehavior of the managed component.
The autonomy manager maintains a knowledge base with static and dynamic entries. The static entries are provided a priori,i.e., at design time. They set up the goals, beliefs, and generic structure of theknowledge base, while the dynamic entries are filled in during operation to capturethe acquired information about concrete situational parameters.
The multicastcommunication primitive makes it possible for the autonomy manager to observethe behavior of the managed component and its interactions with the environmentwithout any probe effect.In its simplest form, the autonomy manager recognizes objects and objectchanges (events) and assigns them to known concepts (see Sect.
2.2.1). It thenselects an action based on event-condition-action rules. If more than one action isapplicable, it uses utility functions to select the action with the highest utility valuefor achieving the desired goals. In a more advanced form, the autonomy manager isbased on a cognitive architecture that supports some form of advanced reasoningand can improve its decision by evaluating past decisions and by the incorporationof learning [Lan09].13.4RFID TechnologyThe easy and fast identification of things is required in many situations, e.g., instores, warehouses, supermarkets etc.
For this purpose, an optical barcode isattached to an object. Reading an optical barcode requires the careful positioningof the object by a human such that a direct line-of-sight between the barcode and thebarcode reader is established.31613 Internet of Things13.4.1 OverviewIn order to be able to automate the process of object identification and eliminate thehuman link, electronic tags (called RFID tags) that can be read from a smalldistance by an RFID reader have been developed. An RFID reader does not requirea direct line-of-sight to the RFID tag.
The RFID tag stores the unique ElectronicProduct Code (EPC) of the attached object. Since an RFID tag has to be attached toevery object, the cost of an RFID tag is a major issue. Due to the standardization ofthe RFID technology by the International Standard Organization (ISO) and themassive deployment of RFID technology, the cost of an RFID tag has been reducedsignificantly over the last few years.The RFID reader can act as a gateway to the Internet and transmit the objectidentity, together with the read-time and the object location (i.e., the location of thereader) to a remote computer system that manages a large database.
![](https://s.studizba.com/z.php?f=/templates/si/i/noavatar.png&w=100&h=100&t=1"){kind=avatar}
