As a mathematical tool, game theory has been used for the analysis of multi-agent systems. Wireless networks are typical examples of such systems, in which communicating nodes access the channel through the CSMA method influencing the other neighboring nodes' access. Different games were examined to model such an environment and investigate its challenging issues. This research reviews different CSMA games presented for wireless MAC and classifies them. Advantages and shortcomings of these games will be recounted and some open research directions for future research, supported.

Edge-Cloud Computing Industrial Internet of Things (ECIIoT) is composed of edge and cloud nodes with Industrial Internet of Things (IIoT) devices to get the Service Function Chain (SFC). The service function chaining placement refers to a series of Virtual Network Functions (VNFs) that are run at edge or cloud nodes in the form of software instances. In the problem of ECIIoT service embedding, the multiple VNFs must be placed for IIoT devices, so how these virtual functions are placed at cloud or edge nodes to minimize the delay is challenging to achieve. In this paper, the placement of virtual functions with considering the edge and cloud nodes is proposed. In our model, the cloud server with edge nodes can run the required functions of IIoT devices in the SFC to decrease the imposed delay and use the computation resource in an efficient way. This is formed as an optimization problem to minimize the delay and residual computing resource consumption and reuse the previous functions. The exact solution of this problem is not available in polynomial time, therefore an efficient approximation algorithm is proposed which solves the problem in three stages. First, it linearizes the non-linear objective function and constraint and approximates them by the convexity of these functions. Then, it solves the relaxed linear problem and finally, it rounds the decision variables in a heuristic way. This solution not only has polynomial time computational complexity but also obtains the near-optimal solution. The simulation results confirm the effectiveness of this approach

Safety and security of Industrial Control Systems (ICS) applied in many critical infrastructures is essential. In these systems, hazards can be due either to system failure or cyber-attacks factors. Accurate hazard detection and reducing reconfiguration time after hazard is one of the most important objectives in these systems. One of the procedures that can reduce the reconfiguration time is determining the cause of hazards and, based on the aforementioned factors, adopting the best commands in reconfiguration time. However, it is difficult to differentiate between different types of hazard because their effects on the system can be similar. With the advent of IoT into ICS, known as IIoT, it has become possible to differentiate the hazards through the adoption of data from different IIoT sensors in the environment. In this article, we propose a risk management approach that identifies hazards based on the physical nature of these systems with the support from the IIoT. The identified hazards fall into four categories: stealthy attack, random attack, transient failure, and permanent failure. Then, the reconfiguration process is run based on the proposed differentiation, which provides a better performance and reconfiguration time. In the experimental section, a fluid storage system is simulated, showing 97% correct differentiation of hazards and reducing in 60% the lost time in the system recovery reconfiguration.