Currently, the network traffic control systems are mainly composed of the Internet core and wired/wireless heterogeneous backbone networks. Recently, these packet-switched systems are experiencing an explosive network traffic growth due to the rapid development of communication technologies. The existing network policies are not sophisticated enough to cope with the continually varying network conditions arising from the tremendous traffic growth. Deep learning, with the recent breakthrough in the machine learning/intelligence area, appears to be a viable approach for the network operators to configure and manage their networks in a more intelligent and autonomous fashion. While deep learning has received a significant research attention in a number of other domains such as computer vision, speech recognition, robotics, and so forth, its applications in network traffic control systems are relatively recent and garnered rather little attention. In this paper, we address this point and indicate the necessity of surveying the scattered works on deep learning applications for various network traffic control aspects. In this vein, we provide an overview of the state-of-the-art deep learning architectures and algorithms relevant to the network traffic control systems. Also, we discuss the deep learning enablers for network systems. In addition, we discuss, in detail, a new use case, i.e., deep learning based intelligent routing. We demonstrate the effectiveness of the deep learning-based routing approach in contrast with the conventional routing strategy. Furthermore, we discuss a number of open research issues, which researchers may find useful in the future.

State-of-the-Art Deep Learning: Evolving Machine Intelligence Toward Tomorrow’s Intelligent Network Traffic Control Systems

The Internet Protocol.

Recent years, Software Defined Routers (SDRs) (programmable routers) have emerged as a viable solution to provide a cost-effective packet processing platform with easy extensibility and programmability Multi-core platforms significantly promote SDRs’ parallel computing capacities, enabling them to adopt artificial intelligent techniques, ie, deep learning, to manage routing paths In this paper, we explore new opportunities in packet processing with deep learning to inexpensively shift the computing needs from rule-based route computation to deep learning based route estimation for high-throughput packet processing Even though deep learning techniques have been extensively exploited in various computing areas, researchers have, to date, not been able to effectively utilize deep learning based route computation for high-speed core networks We envision a supervised deep learning system to construct the routing tables and show how the proposed method can be integrated with programmable routers using both Central Processing Units (CPUs) and Graphics Processing Units (GPUs) We demonstrate how our uniquely characterized input and output traffic patterns can enhance the route computation of the deep learning based SDRs through both analysis and extensive computer simulations In particular, the simulation results demonstrate that our proposal outperforms the benchmark method in terms of delay, throughput, and signaling overhead

Routing or Computing? The Paradigm Shift Towards Intelligent Computer Network Packet Transmission Based on Deep Learning

Network Traffic Matrix (TM) prediction is defined as the problem of estimating future network traffic from the previous and achieved network traffic data. It is widely used in network planning, resource management and network security. Long Short-Term Memory (LSTM) is a specific recurrent neural network (RNN) architecture that is well-suited to learn from experience to classify, process and predict time series with time lags of unknown size. LSTMs have been shown to model temporal sequences and their long-range dependencies more accurately than conventional RNNs. In this paper, we propose a LSTM RNN framework for predicting Traffic Matrix (TM) in large networks. By validating our framework on real-world data from G ´ EANT network, we show that our LSTM models converge quickly and give state of the art TM prediction performance for relatively small sized models.

A Long Short-Term Memory Recurrent Neural Network Framework for Network Traffic Matrix Prediction

This paper presents NeuTM, a framework for network Traffic Matrix (TM) prediction based on Long Short-Term Memory Recurrent Neural Networks (LSTM RNNs). TM prediction is defined as the problem of estimating future network traffic matrix from the previous and achieved network traffic data. It is widely used in network planning, resource management and network security. Long Short-Term Memory (LSTM) is a specific recurrent neural network (RNN) architecture that is well-suited to learn from data and classify or predict time series with time lags of unknown size. LSTMs have been shown to model longrange dependencies more accurately than conventional RNNs. NeuTM is a LSTM RNN-based framework for predicting TM in large networks. By validating our framework on real-world data from GEANT network, we show that our model converges quickly and gives state of the art TM prediction performance.

NeuTM: A neural network-based framework for traffic matrix prediction in SDN

The paper proposes a signaling protocol for network coding-based congestion control and it evaluates its performances The theoretical aspects presented herein justify the expected benefits, i.e. increased throughput of the flows passing through the congested network, as well as the decreased delays and number of packets lost. The signaling protocol proposed was simulated in OMNeT++ to demonstrate its feasibility. The preliminary results are encouraging us to continue the investigations and eventually to combine the mechanism with other techniques that could eliminate congestion, i.e. QoS-aware routing.

Signaling protocol for network coding-based congestion control at network layer

Routing protocols for the Future Internet need to adapt for any given status of the network. However the legacy routing algorithms (e.g. Dijkstra) could not seize the congestion determined by any other event except link failure and/or Layer 2 technology changes. Instead of a sophisticated mechanism we select the most competitive path determination solution to be used by a generic link-state based routing protocol, operating similar to a gearbox system. We suppose the Service Providers work with the Modified Dijkstra's algorithm for a single path routing. Whenever a multi-path delivery is needed, the Ford-Fulkerson scheme might be considered. For Infrastructure Providers the mechanism is similar, except the single path approach, involving this time the Floyd-Warshall algorithm. This paper does not cover the routing protocol implementation, being rather a case study. The major achievement of this proposal refers to a more efficient use of resources (i.e. Available Transfer Rate). The feasibility of this idea was demonstrated with per algorithm OMNET++ simulations on eight-node topology. The scalability is limited to an autonomous system and inter-domain routing solution is for further work.

Case study of a gearbox-like routing algorithm selection in runtime

Whatever the Network of the Future turns out to be, there is little doubt that QoS will constitute a fundamental requirement However, QoS issues and the respective solutions will not remain unchanged New challenges will be raised; new ways of dealing with QoS will be enabled by novel networking concepts and techniques Thus, a fresh approach at the QoS problem will be required This paper addresses QoS in a Future Internet scenario and is focused on three emerging concepts: Network Virtualization, enabling the coexistence of multiple network architectures over a common infrastructure; In-Network Management, improving scalability of management operations by distributing management logic across all nodes; the Generic Path based on the semantic resource management concept, enabling the design of new data transport mechanisms and supporting different types of communications in highly mobile and dynamic network scenarios

Quality of service in the Future Internet

Multi-access and heterogeneous wireless communications are considered to be one of the key drivers in the Future Internet. By using existing technologies, the mobile users will be Always Best Connected (ABC) anytime and anywhere while having full access to new communication services. This paper proposes the system architecture of a connectivity platform designed to implement the ABC paradigm for passengers using public transportation busses. Moreover, the main specifications of entities composing the architecture and the interaction mechanisms between them are also discussed. The key of our approach is the intelligent usage of seamless handover and load balancing techniques between several wireless access technologies offered by multiple telecommunications operators. While keeping the entire mechanism transparent to the data carrier, the system uses efficiently the available resources, maintains service continuity and offers easy and fast deployment. Besides the architectural aspects, the paper discusses some Network State Information management related aspects, presents spotted security issues and proposes a 3G - WLAN vertical handover algorithm based on the functionalities offered by the presented architecture. The experimental evaluation of the proposed algorithm shows a significant increase of the attainable transfer rate compared to classical handover algorithms, better selection of target networks being possible.

Ubiquitous connectivity platform for public transport communication services

This paper presents a possible application of the cross-layer QoS mechanism for network coding in Future Internet. The idea is that the congestion control could be solved for core networks if a network coding-based point-to-multipoint transmission is implemented with quality of service information collected in a bottom-up approach. The implemented testbed combines a real-time demonstrator monitoring transfer rates, one-way delays and BERs at substrate network with network coding simulated in OMNET++ at virtual network level.

http://ieeexplore.ieee.org/iel5/5290637/5306808/05306874.pdf

Andrei Bogdan Rus

Papers

Signaling protocol for network coding-based congestion control at network layer

Case study of a gearbox-like routing algorithm selection in runtime

Quality of service in the Future Internet

Ubiquitous connectivity platform for public transport communication services

Preliminary implementation of point-to-multi-point multicast transmission based on cross-layer QoS and network coding