Neighbor Selection Game in Wireless Ad Hoc Networks

In wireless ad hoc networks, there is no infrastructure to enforce cooperation between nodes. As a result, nodes may act selfishly when running network protocols to conserve their energy resources as much as possible. In this paper, we consider the “neighbor selection” game in which each individual node tries to selfishly choose its neighborhood such that its own energy consumption is optimized. We first focus on a simplified version of this game where nodes know their transmission power before participating in the game. After analyzing the problem, we propose a couple of distributed algorithms to find stable topologies using two kinds of global and local connectivity information. We then take into account the general case where the transmission powers are unknown variables and should be determined during the game. Finally, we evaluate the performance of the proposed algorithms through simulations.     

Joint range assignment and routing to conserve energy in wireless ad hoc networks

In wireless ad hoc networks, energy utilization is perhaps the most important issue, since it corresponds directly to the operational network lifetime. Topology Control (TC) is a well-known energy saving technique which tries to assign transmission ranges of nodes to optimize their energy utilization while keeping the network connected. In current TC schemes, the transmission range of each node is mostly accounted as the exclusive estimator for its energy consumption, while ignoring the amount of data it forwards. Especially when such schemes are coupled with the popular shortest path routing, they usually create a highly-loaded area at the center of the network in which nodes deplete their battery very quickly. In this paper, we introduce efficient strategies that take both load and range into account to handle this problem. We first consider the simple strategy in which a proper transmission range is computed for all nodes of the network to optimize their energy utilization under the presence of the shortest path routing. Inspiring from the results of this strategy, we then propose our combined strategy and argue that a combination of circular paths and shortest paths could result in a much better solution. We also provide detailed analytical models to measure the forwarding load and interference of nodes and then corroborate them with simulation results. Using the combined strategy, the achieved improvement in terms of traffic load, interference, and maximum energy consumption is about 50%, as compared with the simple strategy.     

Energy-efficient topology control in wireless ad hoc networks with selfish nodes

In wireless ad hoc networks there is no fixed infrastructure or centralized controller to enforce cooperation between nodes. Therefore, nodes may act selfishly in running network protocols for conserving their own energy resources. In this paper, we consider the “topology control (TC) game” as the problem of creating an energy-efficient topology in wireless ad hoc networks in the presence of selfish nodes. We define a new TC game in which nodes are able to dynamically adjust their transmission power in a per-packet manner, and try to minimize their energy usage through considering both traffic load and transmission power parameters. After analyzing the problem, we propose several algorithms to find stable topologies in an environment composed of selfish nodes, using two types of global and local connectivity information. Finally, we evaluate the performance of the proposed algorithms by simulations. Our simulation results show that using appropriate local information can interestingly result in more efficient topologies than global information.     

A hybrid machine learning approach for detecting unprecedented DDoS attacks

Service availability plays a vital role on computer networks, against which Distributed Denial of Service (DDoS) attacks are an increasingly growing threat each year. Machine learning (ML) is a promising approach widely used for DDoS detection, which obtains satisfactory results for pre-known attacks. However, they are almost incapable of detecting unknown malicious traffic. This paper proposes a novel method combining both supervised and unsupervised algorithms. First, a clustering algorithm separates the anomalous traffic from the normal data using several flow-based features. Then, using certain statistical measures, a classification algorithm is used to label the clusters. Employing a big data processing framework, we evaluate the proposed method by training on the CICIDS2017 dataset and testing on a different set of attacks provided in the more up-to-date CICDDoS2019. The results demonstrate that the Positive Likelihood Ratio (LR+) of our method is approximately 198% higher than the ML classification algorithms.