Security Framework and Jamming Detection for Internet of Things

The Internet of Things (IoT) consists of billions of people, things and services having the potential to interact with each other and their environment. This highly interconnected global network structure presents new types of challenges from a security, trust and privacy perspective. Hence, security for IoT will be a critical concern that must be addressed in order to enable several current and future applications. The resource constrained devices such as cell phones, PDAs, RFIDs, sensor nodes etc. are the part of IoT. Design process for securing these resource constrained devices is guided by factors like small form factor, good performance, low energy consumption, and robustness to attacks. These design constraints forces us to think of integrating the security features right in to the hardware and software parts of the devices which is also called as embedded security. The research concentrates on embedded security in perspective of software approaches. The IoT system become prone to different security attack, out of all that system is more prone to jamming attack. The goal of research is to design the embedded security framework for IoT and to model the jamming attack and design the defensive technique for Wireless Sensor Network (WSN)-based IoT.
The first part of the thesis proposes the embedded security framework for IoT. The research gives a detailed survey and analysis of embedded security especially in the area of IoT and proposes the security model and threat taxonomy for IoT. The research also highlights the need to provide in-built security in the device itself to provide a flexible infrastructure for dynamic prevention, detection, diagnosis, isolation, and countermeasures against successful breaches. The research proposes the embedded security framework as a feature of software/hardware co-design methodology.
The security framework for IoT also proposes the AES-GCM-based security protocol. The proposed protocol is divided into two components: first is the creation of capability and second component is an application of AES–GCM. AES-GCM is one of the latest authenticated encryption algorithms which provides both message encryption and authentication and can be a good option which will be suited for IoT. AES-GCM core uses a binary Galois Field Multiplier (GFM) for authentication; together with a high-performance AES counter mode cipher to provide high-speed encryption.
The next part of research addresses jamming attack, which is one of the most destructive security attack in the WSN-based IoT. Jamming attack jams the traffic in network by blocking the channel. The different kinds of jamming attack are modelled using unified modelling language (UML). The thesis uses the sequential- and activity- modelling UML approaches to model the behaviour of the jamming attacks. The behavioural modelling and analysis of jamming attack in realistic situations (e.g. sensing in industrial application by following all network rules), gives the clear understanding of jamming attack execution. The research also evaluated the different jamming attack under realistic situations and forms the guidelines to design the countermeasure for jamming attack. The analysis of jamming attack gives the possibility of new kind of jamming attack inside cluster-based network.
The research defines the novel threshold-based countermeasure for reactive jamming attack. The threshold-based jamming countermeasure (TJC) allows the attack into the network and starts its defensive mechanism once it detects the assaults in a network. It uses threshold based mechanism to detect the attack and to cure it. It first detects the jamming node, then informs all neighbouring node about jammer node. The simulation results show that TJC perform in better manner in existence of reactive jamming attack. It demonstrates good performance of TJC by varying traffic interval and number of malicious nodes innetwork. The TJC algorithm is further modified for cluster-based intelligent jamming attack. It also shows good performance under the presence of jamming attack.
The research proposes the game-theory- based countermeasure for detecting different kind of jamming attacks in the network. First, the jamming game is modelled to understand the different moves during attack and non-attack conditions. The game theoretic solution is developed by understanding the game moves. The solution uses the different cross-layer features to design the countermeasures. The proposed detection mechanism shows better energy consumption, throughput, and delay in different realistic situations of network (e.g. varying- amount of traffic and number of malicious nodes) as compared to state-of-art solutions.
The research also contributes in key-management algorithm by proposing cluster-based key management algorithm. The algorithm focused on the management and maintenance of keys under cluster based mobile WSN network. The scheme consider two phases, first for key maintenance which establish the two private keys, home key for own cluster and foreign key when node moves from one cluster to another. The second phase maintain the keys when cluster head (CH) moves from one cluster to another. The proposed algorithm improves the efficiency of key management algorithm in terms of security, mobility, energy efficiency, and scalability of network. The simulation of scheme in different realistic situation shows that proposed solution shows less computational overheads, energy consumption, and delay as compared with state-of-art solution.
The outcome for PhD thesis is proposal for:
- IoT embedded security framework
- IoT threat taxonomy.
- Modelling of jamming attack and proposal for new kind of jamming attack
- Threshold-based countermeasure to detect reactive- and intelligent CH jamming attack.
- Game-theory-based countermeasure for detecting jamming attack by using cross-layer features.
- Efficient key management algorithm for managing the keys under cluster- based mobile WSN network.
In summary, this thesis addresses many important topics of embedded security with special focus on jamming attack detection and defence mechanism and on novel key management for mobile cluster-based WSN. The framework, methods, and techniques proposed in this thesis are, for the most part, applicable to the IoT networks and ubiquitous computing.