A holistic approach to the exploitation of energy harvesting motivated networks: Protocols and countermeasures to DoS attacks
MetadataShow full item record
Internet-of-Things (IoT) and its applications are proliferating, in which a myriad of multi-scale sensors and heterogeneous devices (later in short, nodes) are seamlessly blended for a ubiquitous computing and communication infrastructure. In order to overcome limited battery power and extend the network lifetime, energy harvesting from ambient environment has been increasingly popular and playing an important role in realizing a self-sustainable network. Thus, energy harvesting motivated networks (EHNets) are rapidly emerging and becoming a major building block for diverse IoT applications. In EHNets, a link between two nodes may not be stable due to the variable transmission power levels based on non-uniform energy harvesting rates. Each node is also admittedly vulnerable to Denial-of-Service (DoS) attacks because of the lack of centralized coordination, physical protection, and security requirements in the network protocols. The overall objective of this research is to design, analyze, and evaluate EHNets that can provide reliable, robust, and expected communication performance. We investigate four major research issues. First, light-weight forwarding protocols are proposed to reliably deliver sensory data over time-varying asymmetric links in EHNets. A weighted confirmation scheme, a lazy confirmation scheme, and an asymmetric link aware backoff mechanism are suggested. Second, we propose a light-weight countermeasure to a selective forwarding attack in resource-constrained wireless sensor networks (WSNs), where a randomly selected single checkpoint node is deployed to detect the forwarding misbehaviors. The proposed countermeasure is integrated with timeout and hop-by-hop retransmission techniques to efficiently cover unexpected packet losses due to either forwarding misbehavior or bad channel quality. Third, we propose a new countermeasure, called camouflage-based active detection, to a selective forwarding attack in EHNets. Four adversarial scenarios motivated by energy harvesting and potential forwarding vulnerabilities are also identified and analyzed. Finally, we further extend the camouflage-based active detection to monitor multiple malicious nodes and to detect the forwarding misbehaviors of lurking deep malicious nodes. This countermeasure consists of SlyDog and LazyDog schemes and cooperatively detects the forwarding misbehaviors. The advantages of these techniques are demonstrated through extensive simulation and mathematical analysis.