Research On Software Defined QoS Provisioning Technology Based On Fog Computing In Sensor Networks

With the decrease cost of sensors and the development and popularization of wireless network,Wireless Sensor Networks(WSNs)have been widely used in many areas such as intelligent transportation,environmental monitoring and healthcare.As different applications in WSNs have different Quality of Service(Qo S)requirements for the data sent by the sensor,besides,the energy consumption,computational processing power and network resources of the WSNs are limited and distributed unevenly.Therefore,WSNs require high demands for Qo S.Fog Computing is considered as the next generation of wireless sensor networks due to the flexibility ability to provide computing,storage,and network services.Although fog computing advanced WSNs have solved many Qo S guarantees problems,as data is forwarded by a controller based on content perception in the network,the data can be divided into many classes,and each type of data has different requirements for Qo S.Existing static Qo S provisioning does not work well with content types that are perceived and categorized,so dynamic configuration of Qo S is required.Therefore,the wireless sensor network architecture based on fog computing has put forward new requirements for Qo S provisioning.Many researches have been focused on WSNs based on fog computing.However,there is still no architecture that can well combine the features of fog computing and WSNs.In addition,the dynamic Qo S provisioning mechanism based on fog computing in wireless sensor networks remains a problem to be solved.Based on the above requirements,this paper presents a hierarchical wireless sensor network architecture based on fog computing and a software defined Qo S configuration mechanism is proposed.1)In the WSNs architecture,fog computing is configured at the base station node layer and the sink node layer.Localized data processing is implemented through the fog computing architecture to reduce the delay in network transmission and the load on the network data center.2)The details of software Defined Network(SDN)controller and the details of the fog computing nodes are designed.The Qo S of the WSNs network is ensured through the cooperation of the SDN controller,the base station fog node and the sink fog node.3)Based on Network Calculus(NC),this paper designs a Qo S anomaly prediction mechanism to perceive the emergence of network anomaly in advance through the NC model.Therefore,the network can request Qo S in advance and reallocate the network and computing resources before the network anomaly occurs.4)This paper proposes the software defined dynamic Qo S provisioning mechanism.The design of the flow tables about Qo S configuration in WSNs will be introduced including Qo S request,response,provisioning and so on without the need to make changes to the entire network.5)By setting up a network simulation test environment,the network latency and bandwidth utilization are evaluated.The software defined Qo S provisioning based on fog computing in sensor networks proposed in this paper has been proved to have better scalability and finegrained control ability.