Filters








1,498 Hits in 3.9 sec

Learning Queuing Networks by Recurrent Neural Networks [article]

Giulio Garbi and Emilio Incerto and Mirco Tribastone
2020 pre-print
We encode these equations into a recurrent neural network whose weights can be directly related to model parameters.  ...  We focus on queuing networks, and crucially exploit a deterministic approximation of their average dynamics in terms of a compact system of ordinary differential equations.  ...  Acknowledgements: This work has been partially supported by the PRIN project "SEDUCE" no. 2017TWRCNB.  ... 
doi:10.1145/3358960.3379134 arXiv:2002.10788v1 fatcat:gssoukoidrdijlbwzr5p2explq

Predictive Delay Metric for OLSR Using Neural Networks

Zhihao Guo, Behnam Malakooti
2007 Social Science Research Network  
The key of this mechanism is prediction and evaluation of the mean queuing delay as a routing metric. Neural network methods are used to predict delays.  ...  We investigated the pros and cons of using two types of neural networks, namely Multi-Layer Perceptron (MLP) and Radial Basis Function (RBF), in predicting nonstationary time series (e.g., mean queuing  ...  FUTURE WORK First of all, there is still substantial room for prediction accuracy improvement by using more advanced neural network technologies, such as recurrent neural networks, as they are substantiated  ... 
doi:10.2139/ssrn.2921638 fatcat:qed77bjutzadjk4bnsjc5copmi

Predictive Delay Metric for OLSR Using Neural Networks

Zhihao Guo, Behnam Malakooti
2007 Proceedings of the 3rd International ICSTConference on Wireless Internet  
The key of this mechanism is prediction and evaluation of the mean queuing delay as a routing metric. Neural network methods are used to predict delays.  ...  We investigated the pros and cons of using two types of neural networks, namely Multi-Layer Perceptron (MLP) and Radial Basis Function (RBF), in predicting nonstationary time series (e.g., mean queuing  ...  FUTURE WORK First of all, there is still substantial room for prediction accuracy improvement by using more advanced neural network technologies, such as recurrent neural networks, as they are substantiated  ... 
doi:10.4108/wicon.2007.2140 dblp:conf/wicon/GuoM07 fatcat:nioliyllrzferilnd2pqs5evhm

Echo State Queueing Network: A new reservoir computing learning tool

S. Basterrech, G. Rubino
2013 2013 IEEE 10th Consumer Communications and Networking Conference (CCNC)  
RC models are neural networks which a recurrent part (the reservoir) that does not participate in the learning process, and the rest of the system where no recurrence (no neural circuit) occurs.  ...  ESQNs consist in ESNs where the reservoir has a new dynamics inspired by recurrent RandNNs.  ...  DESCRIPTION OF THE RANDOM NEURAL NETWORK MODEL A Random Neural Network (RandNN) is a specific queuing network proposed in [1] which merges concepts from spiking neural networks and queuing theory.  ... 
doi:10.1109/ccnc.2013.6488435 dblp:conf/ccnc/BasterrechR13 fatcat:oulcwymftfa3vkekygunfl4loq

Deep Learning with Random Neural Networks [chapter]

Erol Gelenbe, Yongha Yin
2017 Lecture Notes in Networks and Systems  
for recurrent networks -Polynomial speed for recurrent gradient descent -Hebbian and reinforcement learning algorithms -Analytical annealing We exploited the analogy with queuing networks, and this also  ...  Excitatory and Inhibitory Spikes (Signals) Inter-neuronal Weights are Replaced by Firing Rates Learning Algorithms for Recurrent Network are O(n 3 ) Multiple Classes (1998) and Multiple Class Learning  ...  CRs use neural networks that essentially form multi-dimensional routing tables that respond immediately to the route performance parameters captured by the packets flowing through them.  ... 
doi:10.1007/978-3-319-56991-8_34 fatcat:u47u4pfj5najvi5ovarm5l7mei

A Review on Delay Prediction Techniques in MANET

Harshita Tuli, Sanjay Kumar
2014 International Journal of Computer Applications  
This paper focusses on different methods adopted by different scientists for estimation and prediction of delay.  ...  So, end to end delay in packet delivery is an important QoS metric in ad-hoc network routing. End-to-End delay comprised of delay involved in transmission, propagation and processing.  ...  In [9] authors have proposed a mobility prediction method based on a recurrent neural network.  ... 
doi:10.5120/18978-0394 fatcat:6uljuhr2wvegnjbvr6mmbpbmee

Graph-neural-network-based delay estimation for communication networks with heterogeneous scheduling policies

Martin Happ, Matthias Herlich, Christian Maier, Jia Lei Du, Peter Dorfinger
2021 ITU Journal  
One of these neural networks is RouteNet, which is based on graph neural networks. However, it is based on simplified assumptions.  ...  This takes neural-network-based delay estimation one step closer to practical use.  ...  by the Federal Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology (BMK) and the Austrian state Salzburg.  ... 
doi:10.52953/tejx5530 fatcat:xtj4p6kcbzhg3lytqeakvplyku

QARC: Video Quality Aware Rate Control for Real-Time Video Streaming via Deep Reinforcement Learning [article]

Tianchi Huang, Rui-Xiao Zhang, Chao Zhou, Lifeng Sun
2018 arXiv   pre-print
Starting from scratch, QARC uses deep reinforcement learning(DRL) algorithm to train a neural network to select future bitrates based on previously observed network status and past video frames, and we  ...  design a neural network to predict future perceptual video quality as a vector for taking the place of the raw picture in the DRL's inputs.  ...  temporarily features via Recurrent Neural Network (RNN).  ... 
arXiv:1805.02482v2 fatcat:z3crzt5eunaarcag6lvddazc4a

Random neural network texture model

Erol Gelenbe, Khaled F. Hussain, Hossam Abdelbaki, Nasser M. Nasrabadi, Aggelos K. Katsaggelos
2000 Applications of Artificial Neural Networks in Image Processing V  
Rubino (INRIA Rennes, France) RNNs in networking Vienna, June 2018 16 / 40 3 -Queuing origins Outline 1 -Random Neural Networks 2 -Perceptual Quality and PSQA 3 -Queuing origins 4 -Extension  ...  Three-layered neural networks: • Input layer • Hidden layer • Output layer a ∈ R Na x ∈ R Nxŷ ∈ R Ny The learning process is restricted to the output weights (readout). • Three sets of neurones, a recurrent  ... 
doi:10.1117/12.382903 fatcat:pcnzvrjnezbodmmpsrljhpnpou

QoS Aware Optimal Confederation based Radio Resource Management Scheme for LTE Networks

2019 International Journal of Engineering and Advanced Technology  
In QOC-RRM scheme we present the hybrid Recurrent Deep Neural Network (RDNN) technique to differentiate the operators by priority wise based on multiple constraints and it control the allocated resource  ...  For routing share queuing criterion data with other schaotic weed optimization (CWO) algorithm are proposed. Once information received each BS schedules the resources for priority user first.  ...  In QOC-RRM scheme we described the hybrid recurrent deep neural network (RDNN) technique used to differentiate the operators by priority wise based on multiple constraints and it controlled the allocated  ... 
doi:10.35940/ijeat.a1014.1291s52019 fatcat:n72g3udi6fczrahdhantjdxllu

Learning Deep Generative Models for Queuing Systems

César Ojeda, Kostadin Cvejoski, Bogdan Georgiev, Christian Bauckhage, Jannis Schuecker, Ramsés J. Sánchez
2021 AAAI Conference on Artificial Intelligence  
We leverage the representation capabilities of Recurrent Marked Point Processes for the temporal dynamics of clients, as well as Wasserstein Generative Adversarial Network techniques, to learn deep generative  ...  To this end, one traditionally relies, within the queuing theory formalism, on parametric analysis and explicit distribution forms.  ...  Acknowledgments The authors of this work were supported by the Fraunhofer Research Center for Machine Learning (RCML) and by the Competence Center for Machine Learning Rhine Ruhr (ML2R), which is funded  ... 
dblp:conf/aaai/OjedaCGBSS21 fatcat:zvtybhb6xvdsjczxblyc2u2lpu

A Dynamic Prediction for Elastic Resource Allocation in Hybrid Cloud Environment

Vipul Chudasama, Madhuri Bhavsar
2020 Scalable Computing : Practice and Experience  
In this paper, we suggest an algorithm using Deep learning and queuing theory concepts that proactively indicate an appropriate number of future computing resources for short term resource demand.  ...  the influence of NNs (Neural Network).  ...  Resource allocation and power management framework is discussed in [23] to predict the workload using long short-term memory (LSTM) recurrent neural network.  ... 
doi:10.12694/scpe.v21i4.1805 fatcat:qko6n6rxjbds5fxtqo3ggzcyri

Neural Networks for Dynamic Shortest Path Routing Problems - A Survey [article]

R. Nallusamy, K. Duraiswamy
2010 arXiv   pre-print
Different shortest path optimization problems can be solved by using various neural networks algorithms.  ...  This paper reviews the overview of the dynamic shortest path routing problem and the various neural networks to solve it.  ...  The performance of the recurrent neural network was demonstrated by Jun Wang for different problems.  ... 
arXiv:0911.2865v2 fatcat:rrwzxh5m2ng2vloksvgeac2tda

Congestion Control Techniques in a Computer Network: A Survey

Mirza WaseemHussain, Sanjay Jamwal, Majid Zaman
2015 International Journal of Computer Applications  
One of the latest approaches to control the congestion is based on Neural Networks is also included in this paper.  ...  In the year 2002 J.Alan Bivens in his paper titled [10] predicts the congestion in a network using neural network learning technique, in which neural networks are used to predict the source of congestion  ...  The recurrent neural networks are trained to regulate the actual size of the queue close to reference value of a queue.  ... 
doi:10.5120/19508-1112 fatcat:nahrmfvgarc6zjxau27cqdvfzu

Delay-Constrained Rate Control for Real-Time Video Streaming with Bounded Neural Network

Tianchi Huang, Rui-Xiao Zhang, Chao Zhou, Lifeng Sun
2018 Proceedings of the 28th ACM SIGMM Workshop on Network and Operating Systems Support for Digital Audio and Video - NOSSDAV '18  
In all considered scenarios, a range based rate control approach outperforms the one without range by 19% to 35% in average QoE improvement.  ...  In this paper, we propose a delay-constrained rate control approach based on end-to-end deep learning.  ...  ., and was part-funded by the National Natural Science Foundation of China under Grant No. 61472204, 61521002, Beijing Key Laboratory of Networked Multimedia (Z161100005016051).  ... 
doi:10.1145/3210445.3210446 dblp:conf/nossdav/HuangZZS18 fatcat:ns3wemnyfre6bg3eapfiv3uety
« Previous Showing results 1 — 15 out of 1,498 results