{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,6]],"date-time":"2026-05-06T16:13:02Z","timestamp":1778083982059,"version":"3.51.4"},"reference-count":49,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T00:00:00Z","timestamp":1669852800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T00:00:00Z","timestamp":1669852800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T00:00:00Z","timestamp":1669852800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-017"},{"start":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T00:00:00Z","timestamp":1669852800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T00:00:00Z","timestamp":1669852800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-012"},{"start":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T00:00:00Z","timestamp":1669852800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2022,12,1]],"date-time":"2022-12-01T00:00:00Z","timestamp":1669852800000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-004"}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Computer Communications"],"published-print":{"date-parts":[[2022,12]]},"DOI":"10.1016\/j.comcom.2022.09.029","type":"journal-article","created":{"date-parts":[[2022,10,6]],"date-time":"2022-10-06T12:15:40Z","timestamp":1665058540000},"page":"184-194","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":182,"special_numbering":"C","title":["Deep reinforcement learning meets graph neural networks: Exploring a routing optimization use case"],"prefix":"10.1016","volume":"196","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3903-6759","authenticated-orcid":false,"given":"Paul","family":"Almasan","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7141-3414","authenticated-orcid":false,"given":"Jos\u00e9","family":"Su\u00e1rez-Varela","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4336-7841","authenticated-orcid":false,"given":"Krzysztof","family":"Rusek","sequence":"additional","affiliation":[]},{"given":"Pere","family":"Barlet-Ros","sequence":"additional","affiliation":[]},{"given":"Albert","family":"Cabellos-Aparicio","sequence":"additional","affiliation":[]}],"member":"78","reference":[{"issue":"4","key":"10.1016\/j.comcom.2022.09.029_b1","doi-asserted-by":"crossref","first-page":"15","DOI":"10.1145\/2829988.2787495","article-title":"A declarative and expressive approach to control forwarding paths in carrier-grade networks","volume":"45","author":"Hartert","year":"2015","journal-title":"ACM SIGCOMM Comput. Commun. Rev."},{"issue":"9","key":"10.1016\/j.comcom.2022.09.029_b2","doi-asserted-by":"crossref","first-page":"1765","DOI":"10.1109\/JSAC.2011.111002","article-title":"The internet topology zoo","volume":"29","author":"Knight","year":"2011","journal-title":"IEEE J. Sel. Areas Commun."},{"key":"10.1016\/j.comcom.2022.09.029_b3","doi-asserted-by":"crossref","first-page":"529","DOI":"10.1038\/nature14236","article-title":"Human-level control through deep reinforcement learning","volume":"518","author":"Mnih","year":"2015","journal-title":"Nature"},{"key":"10.1016\/j.comcom.2022.09.029_b4","series-title":"Mastering chess and shogi by self-play with a general reinforcement learning algorithm","author":"Silver","year":"2017"},{"key":"10.1016\/j.comcom.2022.09.029_b5","series-title":"Why (and how) networks should run themselves","author":"Feamster","year":"2017"},{"issue":"2","key":"10.1016\/j.comcom.2022.09.029_b6","doi-asserted-by":"crossref","first-page":"92","DOI":"10.1109\/MNET.2017.1700200","article-title":"Machine learning for networking: Workflow, advances and opportunities","volume":"32","author":"Wang","year":"2017","journal-title":"IEEE Netw."},{"issue":"3","key":"10.1016\/j.comcom.2022.09.029_b7","doi-asserted-by":"crossref","first-page":"2","DOI":"10.1145\/3138808.3138810","article-title":"Knowledge-defined networking","volume":"47","author":"Mestres","year":"2017","journal-title":"ACM SIGCOMM Comput. Commun. Rev."},{"key":"10.1016\/j.comcom.2022.09.029_b8","doi-asserted-by":"crossref","unstructured":"P. Kalmbach, J. Zerwas, P. Babarczi, A. Blenk, W. Kellerer, S. Schmid, Empowering self-driving networks, in: Proceedings of the Afternoon Workshop on Self-Driving Networks, 2018, pp. 8\u201314.","DOI":"10.1145\/3229584.3229587"},{"key":"10.1016\/j.comcom.2022.09.029_b9","doi-asserted-by":"crossref","unstructured":"A. Valadarsky, M. Schapira, D. Shahaf, A. Tamar, Learning to route, in: Proceedings of the ACM Workshop on Hot Topics in Networks, HotNets, 2017, pp. 185\u2013191.","DOI":"10.1145\/3152434.3152441"},{"issue":"10","key":"10.1016\/j.comcom.2022.09.029_b10","doi-asserted-by":"crossref","first-page":"2260","DOI":"10.1109\/JSAC.2020.3000405","article-title":"RouteNet: Leveraging graph neural networks for network modeling and optimization in SDN","volume":"38","author":"Rusek","year":"2020","journal-title":"IEEE J. Sel. Areas Commun."},{"key":"10.1016\/j.comcom.2022.09.029_b11","doi-asserted-by":"crossref","unstructured":"X. Chen, J. Guo, Z. Zhu, R. Proietti, A. Castro, S.J.B. Yoo, Deep-RMSA: A deep-reinforcement-learning routing, modulation and spectrum assignment agent for elastic optical networks, in: Proceedings of the Optical Fiber Communications Conference, OFC, 2018.","DOI":"10.1364\/OFC.2018.W4F.2"},{"key":"10.1016\/j.comcom.2022.09.029_b12","doi-asserted-by":"crossref","unstructured":"Z. Xu, J. Tang, J. Meng, W. Zhang, Y. Wang, C.H. Liu, D. Yang, Experience-driven networking: A deep reinforcement learning based approach, in: IEEE Conference on Computer Communications, INFOCOM, 2018, pp. 1871\u20131879.","DOI":"10.1109\/INFOCOM.2018.8485853"},{"key":"10.1016\/j.comcom.2022.09.029_b13","doi-asserted-by":"crossref","unstructured":"L. Chen, J. Lingys, K. Chen, F. Liu, Auto: Scaling deep reinforcement learning for datacenter-scale automatic traffic optimization, in: Proceedings of the 2018 Conference of the ACM Special Interest Group on Data Communication, 2018, pp. 191\u2013205.","DOI":"10.1145\/3230543.3230551"},{"key":"10.1016\/j.comcom.2022.09.029_b14","doi-asserted-by":"crossref","unstructured":"A. Mestres, E. Alarc\u00f3n, Y. Ji, A. Cabellos-Aparicio, Understanding the modeling of computer network delays using neural networks, in: Proceedings of the ACM SIGCOMM Workshop on Big Data Analytics and Machine Learning for Data Communication Networks, Big-DAMA, 2018, pp. 46\u201352.","DOI":"10.1145\/3229607.3229613"},{"issue":"11","key":"10.1016\/j.comcom.2022.09.029_b15","doi-asserted-by":"crossref","first-page":"547","DOI":"10.1364\/JOCN.11.000547","article-title":"Routing in optical transport networks with deep reinforcement learning","volume":"11","author":"Su\u00e1rez-Varela","year":"2019","journal-title":"IEEE\/OSA J. Opt. Commun. Networking"},{"key":"10.1016\/j.comcom.2022.09.029_b16","series-title":"Relational inductive biases, deep learning, and graph networks","author":"Battaglia","year":"2018"},{"issue":"1","key":"10.1016\/j.comcom.2022.09.029_b17","doi-asserted-by":"crossref","first-page":"61","DOI":"10.1109\/TNN.2008.2005605","article-title":"The graph neural network model","volume":"20","author":"Scarselli","year":"2008","journal-title":"IEEE Trans. Neural Netw."},{"key":"10.1016\/j.comcom.2022.09.029_b18","unstructured":"J. Gilmer, S.S. Schoenholz, P.F. Riley, O. Vinyals, G.E. Dahl, Neural message passing for quantum chemistry, in: Proceedings of the International Conference on Machine Learning- Vol. 70, ICML, 2017, pp. 1263\u20131272."},{"key":"10.1016\/j.comcom.2022.09.029_b19","unstructured":"https:\/\/github.com\/knowledgedefinednetworking\/DRL-GNN."},{"key":"10.1016\/j.comcom.2022.09.029_b20","series-title":"Gated graph sequence neural networks","author":"Li","year":"2015"},{"key":"10.1016\/j.comcom.2022.09.029_b21","series-title":"Graph attention networks","author":"Veli\u010dkovi\u0107","year":"2017"},{"key":"10.1016\/j.comcom.2022.09.029_b22","unstructured":"P.W. Battaglia, R. Pascanu, M. Lai, D.J. Rezende, et al., Interaction networks for learning about objects, relations and physics, in: Proceedings of Advances in Neural Information Processing Systems, NIPS, 2016, pp. 4502\u20134510."},{"issue":"3\u20134","key":"10.1016\/j.comcom.2022.09.029_b23","first-page":"279","article-title":"Q-learning","volume":"8","author":"Watkins","year":"1992","journal-title":"Mach. Learn."},{"key":"10.1016\/j.comcom.2022.09.029_b24","series-title":"Playing atari with deep reinforcement learning","author":"Mnih","year":"2013"},{"key":"10.1016\/j.comcom.2022.09.029_b25","doi-asserted-by":"crossref","unstructured":"J. Kuri, N. Puech, M. Gagnaire, Diverse routing of scheduled lightpath demands in an optical transport network, in: Proceedings of the IEEE International Workshop on Design of Reliable Communication Networks, DRCN, 2003, pp. 69\u201376.","DOI":"10.1109\/DRCN.2003.1275340"},{"key":"10.1016\/j.comcom.2022.09.029_b26","series-title":"ITU-T Recommendation g.709\/y.1331: Interface for the optical transport network","year":"2019"},{"key":"10.1016\/j.comcom.2022.09.029_b27","series-title":"Reinforcement Learning: An Introduction","author":"Sutton","year":"2018"},{"key":"10.1016\/j.comcom.2022.09.029_b28","series-title":"On the properties of neural machine translation: Encoder-decoder approaches","author":"Cho","year":"2014"},{"key":"10.1016\/j.comcom.2022.09.029_b29","unstructured":"M. Abadi, P. Barham, J. Chen, Z. Chen, A. Davis, J. Dean, M. Devin, S. Ghemawat, G. Irving, M. Isard, et al., Tensorflow: A system for large-scale machine learning, in: Proceedings of the 12th USENIX Symposium on Operating Systems Design and Implementation, OSDI, 2016, pp. 265\u2013283."},{"key":"10.1016\/j.comcom.2022.09.029_b30","series-title":"Openai gym","author":"Brockman","year":"2016"},{"key":"10.1016\/j.comcom.2022.09.029_b31","doi-asserted-by":"crossref","unstructured":"L. Bottou, Large-scale machine learning with stochastic gradient descent, in: Proceedings of the International Conference on Computational Statistics, COMPSTAT, 2010, pp. 177\u2013186.","DOI":"10.1007\/978-3-7908-2604-3_16"},{"issue":"5","key":"10.1016\/j.comcom.2022.09.029_b32","doi-asserted-by":"crossref","first-page":"363","DOI":"10.1364\/JON.3.000363","article-title":"Wavelength converter placement in least-load-routing-based optical networks using genetic algorithms","volume":"3","author":"Hei","year":"2004","journal-title":"J. Opt. Netw."},{"key":"10.1016\/j.comcom.2022.09.029_b33","series-title":"Fast emergency paths schema to overcome transient link failures in OSPF routing","author":"Barreto","year":"2012"},{"issue":"3","key":"10.1016\/j.comcom.2022.09.029_b34","doi-asserted-by":"crossref","first-page":"35","DOI":"10.1145\/1070873.1070877","article-title":"Achieving sub-second IGP convergence in large IP networks","volume":"35","author":"Francois","year":"2005","journal-title":"ACM SIGCOMM Comput. Commun. Rev."},{"issue":"4","key":"10.1016\/j.comcom.2022.09.029_b35","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1145\/2534169.2486019","article-title":"B4: Experience with a globally-deployed software defined WAN","volume":"43","author":"Jain","year":"2013","journal-title":"ACM SIGCOMM Comput. Commun. Rev."},{"key":"10.1016\/j.comcom.2022.09.029_b36","series-title":"Exploring Network Structure, Dynamics, and Function Using Networkx","author":"Hagberg","year":"2008"},{"key":"10.1016\/j.comcom.2022.09.029_b37","series-title":"How powerful are graph neural networks?","author":"Xu","year":"2018"},{"issue":"8","key":"10.1016\/j.comcom.2022.09.029_b38","doi-asserted-by":"crossref","first-page":"836","DOI":"10.1364\/JOCN.5.000836","article-title":"Efficient resource allocation for all-optical multicasting over spectrum-sliced elastic optical networks","volume":"5","author":"Gong","year":"2013","journal-title":"IEEE\/OSA J. Opt. Commun. Networking"},{"issue":"9","key":"10.1016\/j.comcom.2022.09.029_b39","doi-asserted-by":"crossref","first-page":"1520","DOI":"10.1109\/LCOMM.2012.070512.120740","article-title":"A two-population based evolutionary approach for optimizing routing, modulation and spectrum assignments (RMSA) in O-OFDM networks","volume":"16","author":"Gong","year":"2012","journal-title":"IEEE Commun. Lett."},{"issue":"1","key":"10.1016\/j.comcom.2022.09.029_b40","doi-asserted-by":"crossref","first-page":"26","DOI":"10.1109\/JSAC.2013.130104","article-title":"Elastic spectrum allocation for time-varying traffic in flexgrid optical networks","volume":"31","author":"Klinkowski","year":"2012","journal-title":"IEEE J. Sel. Areas Commun."},{"key":"10.1016\/j.comcom.2022.09.029_b41","series-title":"IEEE International Conference on Computer Communications","first-page":"1503","article-title":"A study of the routing and spectrum allocation in spectrum-sliced elastic optical path networks","author":"Wang","year":"2011"},{"issue":"9","key":"10.1016\/j.comcom.2022.09.029_b42","doi-asserted-by":"crossref","first-page":"1354","DOI":"10.1109\/JLT.2011.2125777","article-title":"Elastic bandwidth allocation in flexible OFDM-based optical networks","volume":"29","author":"Christodoulopoulos","year":"2011","journal-title":"J. Lightwave Technol."},{"key":"10.1016\/j.comcom.2022.09.029_b43","doi-asserted-by":"crossref","DOI":"10.1016\/j.comnet.2021.107891","article-title":"Scaledrl: A scalable deep reinforcement learning approach for traffic engineering in SDN with pinning control","volume":"190","author":"Sun","year":"2021","journal-title":"Comput. Netw."},{"issue":"10","key":"10.1016\/j.comcom.2022.09.029_b44","doi-asserted-by":"crossref","first-page":"2249","DOI":"10.1109\/JSAC.2020.3000371","article-title":"CFR-RL: Traffic engineering with reinforcement learning in SDN","volume":"38","author":"Zhang","year":"2020","journal-title":"IEEE J. Sel. Areas Commun."},{"issue":"7","key":"10.1016\/j.comcom.2022.09.029_b45","doi-asserted-by":"crossref","first-page":"2198","DOI":"10.1109\/JSAC.2020.3041385","article-title":"On deep reinforcement learning for traffic engineering in SD-WAN","volume":"39","author":"Troia","year":"2020","journal-title":"IEEE J. Sel. Areas Commun."},{"key":"10.1016\/j.comcom.2022.09.029_b46","doi-asserted-by":"crossref","unstructured":"F. Geyer, G. Carle, Learning and generating distributed routing protocols using graph-based deep learning, in: Proceedings of the ACM SIGCOMM Workshop on Big Data Analytics and Machine Learning for Data Communication Networks, Big-DAMA, 2018, pp. 40\u201345.","DOI":"10.1145\/3229607.3229610"},{"key":"10.1016\/j.comcom.2022.09.029_b47","doi-asserted-by":"crossref","unstructured":"H. Zhu, V. Gupta, S.S. Ahuja, Y. Tian, Y. Zhang, X. Jin, Network planning with deep reinforcement learning, in: Proceedings of the 2021 ACM SIGCOMM 2021 Conference, 2021, pp. 258\u2013271.","DOI":"10.1145\/3452296.3472902"},{"key":"10.1016\/j.comcom.2022.09.029_b48","series-title":"2021 IEEE 29th International Conference on Network Protocols","first-page":"1","article-title":"Is machine learning ready for traffic engineering optimization?","author":"Bern\u00e1rdez","year":"2021"},{"key":"10.1016\/j.comcom.2022.09.029_b49","doi-asserted-by":"crossref","unstructured":"H. Mao, M. Schwarzkopf, S.B. Venkatakrishnan, Z. Meng, M. Alizadeh, Learning scheduling algorithms for data processing clusters, in: Proceedings of ACM SIGCOMM, 2019, pp. 270\u2013288.","DOI":"10.1145\/3341302.3342080"}],"container-title":["Computer Communications"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0140366422003784?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0140366422003784?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2025,9,25]],"date-time":"2025-09-25T23:38:59Z","timestamp":1758843539000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S0140366422003784"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,12]]},"references-count":49,"alternative-id":["S0140366422003784"],"URL":"https:\/\/doi.org\/10.1016\/j.comcom.2022.09.029","relation":{},"ISSN":["0140-3664"],"issn-type":[{"value":"0140-3664","type":"print"}],"subject":[],"published":{"date-parts":[[2022,12]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"Deep reinforcement learning meets graph neural networks: Exploring a routing optimization use case","name":"articletitle","label":"Article Title"},{"value":"Computer Communications","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.comcom.2022.09.029","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2022 Elsevier B.V. All rights reserved.","name":"copyright","label":"Copyright"}]}}