بکار گیری هوش مصنوعی در بهبود شبکه های حسگر بی سیم قسمت اول
Artiﬁcial Intelligence for Wireless Sensor Networks Enhancement
Whereas the main objective of Artiﬁcial ntellience is to develop systems that emulate the intellectual and interaction abilities of a human bein the Distributed Artiﬁcial ntellience pursues the same objective but focusin on human bein societies O’are et al, 2 A paradim in current use for the development of Distributed Artiﬁcial ntellience is based on the notion of multiaent systems A multiaent system is formed by a number of interactin intellient systems called aents, and can be implemented as a softare proram, as a ded icated computer, or as a robot Russell orvin, 2 ntellient aents in a multiaent system interact amon each other to oranie their structure, assin tass, and interchane nolede
Concepts related to multiaent systems, artiﬁcial societies, and simulated oraniations, cre ate a ne and risin paradim in computin hich involves issues as cooperation and compe tition, coordination, collaboration, communication and lanuae protocols, neotiation, con sensus development, conﬂict detection and resolution, collective intellience activities con ducted by aents e problem resolution, plannin, learnin, and decision main in a dis tributed manner, conitive multiple intellience activities, social and dynamic structurin, decentralied administration and control, safety, reliability, and robustness service uality parameters
Distributed intellient sensor netors can be seen from the perspective of a system com posed by multiple aents sensor nodes, ith sensors orin amon themselves and form in a collective system hich function is to collect data from physical variables of systems hus, sensor netors can be seen as multiaent systems or as artiﬁcial oranied societies that can perceive their environment throuh sensors
ut, the uestion is ho to implement Artiﬁcial ntellience mechanisms ithin Wireless en sor etors W s here are to possible approaches to the problem accordin to the ﬁrst approach, desiners have in mind the lobal objective to be accomplished and desin both, the aents and the interaction mechanism of the multiaent system n the second approach, the desiner conceives and constructs a set of selfinterested aents hose then evolve and interact in a stable manner, in their structure, throuh evolutionary techniues for learnin he same difﬁculty applies hen orin ith a W perspective seen from the
perspective of DAI. Can the principles, algorithms and application of Distributed Artiﬁcial Intelligence be used to optimize a network of distributed wireless sensors? Is it possible to implement a solution that enables a sensor network to behave as an intelligent multi-agent sstem? rom a perspective of multi-agents, artiﬁcial societies, and simulated organizations, how must a distributed sensor network be installed in an efﬁcient manner and achieve the proposed obectives of taking measures of phsical variables b itself? hat are the union points between Distributed Artiﬁcial Intelligence and ireless sensor networks? he fundamental idea is this chapter is to propose a model that enables a highl distributed sensor network to behave intelligentl as a multi-agent sstem.
2. Wireless Sensor Networks
A ensor etwork is a sstem that consists of thousands of ver small stations called sensor nodes. he main function of sensor nodes it is to monitor, record and notif a speciﬁc condition at various locations to other stations. Also, a is a group of specialized transducers with a communications infrastructure intended to monitor and record conditions at diverse locations. Commonl monitored parameters are temperature, humidit, pressure, wind direction and speed, illumination intensit, vibration intensit, sound intensit, power-line voltage, chemical concentrations, pollutant levels and vital bod functions.
ensor nodes can be imagined as small computers, etremel basic in terms of their interfaces and their components. Although these devices have a ver little capabilit on their own the have substantial processing capabilities when the are working as an aggregate, C et al., . ach node in a sensor network is tpicall euipped with a radio transceiver or other wireless communications device, a small microcontroller, and an energ source, usuall a batter. A sensor node might var in size from that of a shoebo down to the size of a grain of dust omer attern, . A sensor network normall constitutes a wireless ad-hoc network, meaning that each sensor supports a multi-hop routing algorithm several nodes ma forward data packets to the base station. It is important to underline that s are subect to more severe power constraints than DAs, mobile phones, or laptops. he whole network is usuall under the administration of one controller the base station. he main functionalit of the base station is to act as gatewa to another network, and is a powerful data processor and storage center. Advances in microelectronics and wireless communications have made s the predict panacea for attacking a host of large-scale decision and information processing tasks. he applications for s are varied, tpicall involving some kind of monitoring, tracking, or controlling. peciﬁc applications include habitat monitoring, obect tracking, nuclear reactor control, ﬁre detection, and trafﬁc monitoring. In a tpical application, a is scattered in a region where it is meant to collect data through its sensor nodes. A number of s have been deploed for environmental monitoring Davoudani et al., . an of these have been short lived, often due to the prototpe nature of the proects. ireless sensor networks have been developed for machiner Condition-ased aintenance C since the offer signiﬁcant cost savings and enable new functionalities.
Although a number of new sstems and technologies have been developed, a number of new problems or challenges are et to be solved or improved on. amples of such problems are optimal routing strategies, lifespan of the , lifetime of the nodes are often ver limited, reconﬁgurabilit without redeploment, etc.
inall, since s become popular there is not a common platform. ome representative designs have broader users and developer communities, such as erkele otes, which was