In ad hoc networsk, nodes operating under dynamic topology are often correlated in their behavior. Correlated behavior may poses devastating impact towards network survivability. It can be investigated by observing node activities such as forwarding and malicious activity. To assess systematic network survivability in providing efficient communication in the events of correlated behaviors, it requires a stochastic abstraction of node behavior to capture correlated event. We proposed in this paper, a novel Semi Markov with epidemic model for modeling correlated node behavior in mobile ad hoc networks. The model consists of two stages. First stage characterizes node behavior and its transition based on Semi Markov process while at second stage correlated degree is proposed using epidemic model to generated correlated behavior event sequence. The model is able to predict correlated degree of nodes from the current status of neighboring nodes. This criteria is important in understanding the spread of node particularly for misbehave node and in developing counter measures for survival network.

In wireless sensor networks, nodes operating under dynamic topology are often correlated with their behavior. Correlated behavior may pose devastating impact towards network connectivity. A node may change its behaviour from cooperative node to misbehave node which directly affects the network’s connectivity. Misbehaviour nodes tend to have correlated effect which creates partitioning within the network. To improve network connectivity in providing an efficient communication in the events of the correlated behaviors, a new formulation of correlated degree to perform network clustering is required. This paper proposes a formulation on correlated degree using 3D Euclidean distance to achieve higher network connectivity under correlated node behavior. The key idea behind the 3D Euclidean distance in network clustering is to identify a set of sensors whose sensed values present some data correlation referring to correlated degree. The correlated degree is formulated based on three-point distance within a correlation region to identify the level of node correlation within neighboring nodes. In addition, the correlated degree also be able to detect the same group of node behavior which is grouped in correlated regions. 3D Euclidean distance is shown in mathematical analysis and how the new formulation calculates correlated degree is also discussed. It is also expected that the new 3D Euclidean distance formulation may help correlation region to change it cluster formation dynamically to achieve the required network connectivity.

This paper presents new model to study the survivability of ad hoc network in which node behaviors are statistically correlated. The model tries to identify explicitly the correlated node events that cause network to isolate. The model characterizing the node behavior transition based on Semi Markov process and predicts how nodes affected neighboring activity to which it is important for understanding their potential damages, and for developing countermeasures to secure and ensure survivable of mobile ad hoc network. For survivability analysis, the model introduces correlated degree as a new function of survivability to measure nodes connectivity. The study evaluates the impact of correlated node behavior particularly selfish, malicious and fails nodes toward network resilience and survivability. The results show that correlated node behaviors have more adverse effects on the survivability.

This paper present a new survivability model based on correlated node behavior to study the global survivability of mobile ad hoc networks. The model uses k-correlated survivability model for correlated node behavior of mobile ad hoc networks which is an extended version of k-connectivity of individual node model. The k-correlated survivability model takes correlated degree to measure node correlation as a new function of survivability. The study evaluates the impact of correlated node behavior particularly selfish, malicious and fails nodes toward network resilience and survivability. The results show that correlated node behaviors have more adverse effects on the survivability.

Network survivability analysis in MANETs was hardly an issue in the early years of wireless technology because there were no critical network system that depended on wireless technology yet. Today, network survivability analysis is an essential aspect of reliable communication especially in MANETs. Although various methods have been proposed to measure network survivability analysis in MANETs, no related review has been published as to date for this topic. Thus, a comprehensive review of this body of work would be beneficial to researchers to have an overview of the current state of research trend in this area. This paper provides a systematic literature review (SLR) of the state of the art approach in network survivability analysis in MANETs. We used studies from a number of relevant article sources, and our results showed the existence of twenty six (26) articles. From this SLR we found that the existing of analysis method is focusing on individual node in which the node is treated as independent event. Furthermore, the analysis also reveals the less popular methods in analyzing network survivability are with statistical methods such as regression analysis and survival analysis. The implication of this study is to give a clear direction to future researchers in this area for a better and accurate analysis in measuring network survivability in MANETs. (C) 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).