Abstract:Recent years have witnessed the deep integration of embedded systems and wireless networked systems such as computational RFIDs and embedded sensor networks.These systems comprise of tiny embedded chips,low-power radios,and various sensors.They allow for an instrumentation of the physical world at an unprecedented scale and density,thus enabling a new generation of monitoring and control applications.Traditionally,these systems are highly application customized because of their resource constraints.Customized systems have the benefits of easy developments and good performance.They are,however,not a good fit for wide deployments in the future.If these systems were highly application dependent and inflexible to upgrade,our technical innovations will be severely hindered since the replacement of a large number of sensor products take a long time.This situation would be similar to today's Internet where the deployment of a new protocol is not easy.This calls for a software defined architecture.Software-defined network has been a hot research topic for the networking community.Software design for sensor networks has many differences from the Internet.Most importantly,sensor networks are primarily used for information gathering while the Internet is primarily used for information transportation.Moreover,sensors are characterized as resource constrained due to the small form factor,limited battery supply,and low prices.In this article,we focus on the design of software-defined architecture for wireless sensor networks.The technology of traditional software-defined network mainly focuses on how to make the routing behavior software-defined.Software-defined sensor networks focus on not only software-defined routing,but also other aspects such as software-defined sensing.For the design of a generic and efficient softwaredefined sensor network architecture,we first examine some major challenges,e.g.,the design of a generic software-defined architecture,efficient allocation of resources,rapid update and real-time control,maintenance of network consistency,etc.Among these challenges,the design of a generic software-defined architecture for sensor networks is especially important,and it is the basis for controlling the network in a flexible way.The efficient allocation of resources,rapid update and real-time control,and maintenance of network consistency are also key challenges in implementing software-defined sensor network.The challenge of security already exists in traditional sensor networks as well as the Internet.However,how to design a secure and lightweight security mechanism in software-defined sensor network faces new challenges.To address part of the above challenges,we review some useful techniques.Then,we also present a novel taxonomy for software-defined sensor network according to different abstractions of functionalities,including software-defined sensing,software-defined routing,software-defined measurements,software-defined debugging,and others.The purpose of this article is to summarize the current designs in software defined sensor network,so as to explore a larger research space in this direction.In the end,we also summarize some possible future directions,e.g.,the possible combination of wireless reprogramming and remote debugging,and the design of agile programming model.In the future,we believe that a well-designed software-defined sensor network architecture can greatly facilitate software development for deployed networks,allowing rapid technical innovations.

A survey on Machine Learning Software-Defined Wireless Sensor Networks (ML-SDWSNs): Current status and major challenges

Towards a Software-Defined Architecture for Wireless Sensor Networks

A Survey of Machine Learning for Information Processing and Networking

Next-Generation WSN for Environmental Monitoring Employing Big Data Analytics, Machine Learning and Artificial Intelligence

A Survey of Machine Learning in Wireless Sensor Netoworks from Networking and Application Perspectives

A Survey Of Machine Learning In Wireless Sensor Netoworks

A Survey of Networking Applications Applying the Software Defined Networking Concept Based on Machine Learning

The Intersection of Machine Learning and Wireless Sensor Network Security for Cyber-Attack Detection: A Detailed Analysis

A Survey of Machine Learning Techniques Applied to Software Defined Networking (SDN): Research Issues and Challenges

Software Defined Networks in Wireless Sensor Architectures

Deep Learning Challenges and Prospects in Wireless Sensor Network Deployment

Wireless for Machine Learning

Survey on SDN based network intrusion detection system using machine learning approaches

Machine Learning in Wireless Sensor Networks: Algorithms, Strategies, and Applications

Wireless Sensor Networks anomaly detection using Machine Learning: A Survey

Wireless sensor networks and machine learning centric resource management schemes: A survey

Evolution of Software-Defined Sensor Networks

Software-Defined Architectures and Technologies for Underwater Wireless Sensor Networks: A Survey.

A survey on wireless multimedia sensor networks

Design and Implementation of a Novel Software-Defined Wireless Sensor Network

Introducing Machine Learning to Wireless Sensor Networks