Research Of IoT Service Discovery Algorithm Based On WSN

The internet of things set off the third wave of the information technology industry,drawing the extensive attention from academia and industry.The Internet of things is made up of large-scale heterogeneous devices and networks,so the heterogeneity and scalability are important problems facing the Internet of things.SOA(Service-Oriented Architecture)encapsulates device functions and middlevware into services,providing unified service interfaces,shielding the heterogeneity of the underlying devices and networks,reducing the coupling between service providers and service users,and improving scalability.Therefore,SOA is a commonly used architecture of IOT.In addition,IoT is a global network composed of networks.There are many IoT solutions based on WSN,which is an importent component,key enable technology of IoT.And SOA has been successfully applied in WSN with the development of WSN Technology,sensing as service model,and multi-sense architecture.The SOA architecture of the IoT encapsulates the capability of the device as a service,performs service management and computing,and provides a service interface for the application.SOA's service computing includes service discovery,service selection,service interruption recovery,service composition and so on.Service discovery is the foundation of other services computing,which is of great significance.This paper presents a IoT service discovery algorithm based on WSN.The Internet of things contains many resource constrained devices.In order to improve the lifetime and scalability of the network and reduce energy consumption,we first use clustering technology to establish the network organization structure,and then propose a distributed service discovery algorithm based on cluster structure.At last,the introduced algorithm will be integrated in an Open-IoT platform.The main contents of this paper are as follows:1.SOA oriented cluster algorithm:combining the computing needs of service computing and edge computing in WSN,we decompose the function of cluster head in traditional clustering algorithm to transmit nodes and compute nodes,so as to provide foundation for service discovery algorithm based on cluster.The problems of hotspot and connectivity problems of transmit nodes are solved by controlling the distribution density and position.The distribution density of the compute nodes is controlled to balance the network load and alleviate the energy hole problem.2.Service discovery algorithm employing cluster technology:clustering techniques are used to organize nodes,and cluster heads are used as storage and computing nodes for service discovery.Using Bloom filter technology to generate binary encoding of service description,in order to save storage space required for the service description.The encodings are stored in distributed cluster heads based on similarity,and the service discovery request is relayed by the same way.3.Design scheme of IoT service discovery algorithm based on WSN integrated in the Open-IoT platform:WSN is connected to the Open-IoT platform as a virtual device to register the services provided by WSN.When the requested service is provided by WSN,the platform forwards the service discovery request to WSN,runs the distributed service discovery algorithm,calculates the load of the sensor nodes supporting the service,and selects the node with least load as the response node of the service discovery request.The simulation results show that the proposed algorithm has a better effect compared with the existing algorithm.SOA oriented clustering algorithm can increase the lifetime of network,reduce packet loss rate and improve throughput when energy consumption of service discovery increases.The service discovery algorithm using clustering technology can improve the success rate of service discovery,shorten the response time,and reduce the energy consumption.