International audienceWireless Mesh Networks (WMNs) are ad-hoc networks where no infrastructure exists and where some of the participating nodes can also act as routers to forward traffic on behalf of others. WMNs have been proposed for various applications and can greatly vary in terms of size (from a handful of nodes to thousands) and complexity (static or mobile nodes, scarce or dense etc.). Efficient and robust IP routing is essential for these networks and its performance can greatly affect all other network performance metrics such as throughput, delay, packet delivery ratio etc. A lot of routing algorithms have been proposed ever since the advent of WMNs and research is still ongoing on the topic. All of the proposed solutions are based on the principle of distributed routing where routers exchange signalization messages in order to discover the network topology. In the present paper we propose a centralized routing based on the relatively recent paradigm of Software Defined Networking (SDN). We have used NS3 to simulate a dynamic mesh network where a subset of nodes are mobile and a source host generates traffic destined to a sink host. We evaluated four common network performance metrics namely packet delivery ratio, packet drop ratio due to route unavailability, average throughput and signalling overhead. By making appropriate assumptions, we compared our solution with three of the most widely used routing protocols for WMNs namely DSDV, AODV and OLSR. Our simulations demonstrate that, under certain conditions, our solution outperforms the aforementioned distributed routing algorithms and can improve network performance
Predicting the bandwidth utilization on network links can be extremely useful for detecting congestion in order to correct them before they occur. In this paper, we present a solution to predict the bandwidth utilization between different network links with a very high accuracy. A simulated network is created to collect data related to the performance of the network links on every interface. These data are processed and expanded with feature engineering in order to create a training set. We evaluate and compare three types of machine learning algorithms, namely ARIMA (AutoRegressive Integrated Moving Average), MLP (Multi Layer Perceptron) and LSTM (Long Short-Term Memory), in order to predict the future bandwidth consumption. The LSTM outperforms ARIMA and MLP with very accurate predictions, rarely exceeding a 3% error (40% for ARIMA and 20% for the MLP). We then show that the proposed solution can be used in real time with a reaction managed by a Software-Defined Networking (SDN) platform.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.