Iraj Saniee

Iraj Saniee is Head of the Mathematics & Algorithms Research Group at Nokia Bell Labs, in Murray Hill, NJ. The Group consists of multiple departments focusing on Communication, Systems, Networks and Machine Learning in Murray Hill (USA), Cambridge (United Kingdom), and Paris-Saclay (France). The departments are headed by Carl Nuzman, Armen Aghasaryan, Gerard Burnside, Joachim Wabnig  and Chun-Nam Yu.

Iraj's research interests span a spectrum of topics in applied mathematics, modeling & analysis and optimization of communication networks and systems with focus on the structure and architecture of networks, scalable algorithms for network problems, application to design tools and control mechanisms for emerging systems and more recently foundations and applications of machine learning.

Prior to Bell Labs and until 1998, he was a research director in the Information Sciences Lab at Bellcore in Morristown, New Jersey.

In addition to his research in support of Nokia products and services, in the past decade Iraj has been the recipient of DARPA, AFOSR and NIST contracts and grants on such topics as control mechanisms for MANETs, large-scale geometry of networks and inference from large graphs. His current focus is on deep learning, large graph analytics and distributed self-organizing network control.

I work in Algorithms, driven by mathematics and sound foundations, and motivated by problems that arise in communication.  This is diffferent from finding a quick recipe for an instance of a problem using a good guess.  The former aims for broad generalizability and the latter for immediate utility.  Both approaches have had spectacular successes in the past 70 years since programable computing machines came into being.  The simplex algorithm and the Metropolis algorithm are examples of the former and rule-based systems are examples of the latter.  The two approaches are different and represent essentially different views of the world. 

I also (try to) follow John Tukey's dictum: An approximate answer to the right question (which is often vague) is far better than an exact answer to the wrong question (which can always be made precise).

Ph. D. degree in operations research and control theory,  University of Cambridge
M. A. (Hon.) degree in mathematics,  University of Cambridge
B. A. (Hon.) degree in mathematics,  University of Cambridge

Associate Editor of IEEE Transaction of Information Theory

Serves on program committees of numerous ACM, IEEE and INFORMS conferences and workshops including NeurIPS, ICML, WebConf, KDD, WSDM, and Infocom

Has served as reviewer for NSF, NSF of Ireland, AFOSR, INRIA and other research granting agencies

Member/past member of IFIP WG 7.3, IEEE, and INFORMS 

Former member of the American Mathematical Society, Society for Industrial and Applied Mathematics

Past Chair of the Telecommunication Section of INFORMS

Formerly on the Editorial Board of Operations Research

An analytical solution to the multicommodity network fow problem with weighted random routing 
O. Narayan, Iraj Saniee, Applied Network Science, Springer Open Access, June 16, 2021.

Efficient Deep Approximation of GMMs
S. Jalali, C. Nuzman, Iraj Saniee.  Proceedings of 33rd NeurIPS Conference, December 2019 (see also the summary poster)

Digital innovation networks at the nexus of new productivity growth
S. Kamat, S. Prakash, Iraj Saniee, M. Weldon.  To appear in BLTJ.

Efficient Deep Learning of GMMs
S. Jalali, C. Nuzman, Iraj Saniee, arXiv, February 2019

Fast approximation algorithms for p-centres in large δ-hyperbolic graphs 
K. Edwards, S. Kennedy, Iraj Saniee, Algorithmica, Vol 80, Issue 12, pp 3889-3907,  Dec 2018, https://doi.org/10.1007/s00453-018-0425-6

Will productivity growth return in the new digital era?
Iraj Saniee, S. Kamas, S. Prakash, M. Weldon, Bell Labs Technical Journal, Spring 2017

Quantifying the benefits of infrastructure sharing 
M. Andrews, M. Bradonjic, Iraj Saniee, NetEcon '17, Proc. 12th Workshop on Economics of Networks, Systems, MIT, 2017

Fast approximation algorithms for p-centres in large δ-hyperbolic graphs 
K. Edwards, S. Kennedy, Iraj Saniee, Proceedings of WAW16, Springer, 2016. 

Scalable algorithms for large and dynamic networks: Reducing big data for small computations 
Iraj Saniee, Bell Labs Technical Journal, Vol: 20, pp 23-33, Wiley, July 2015

Lack of hyperbolicity in asymptotic Erdos-Renyi random graphs
O. Narayan, Iraj Saniee, G. H. Tucci, Internet Mathematics, Vol. 11-3, Taylor & Francis, May 2015.

Bootstrap percolation in periodic trees
M. Bradonjic, Iraj Saniee, ANALCO15, SIAM, January 2015.

Non-concave utility maximization in locally coupled systems with applications to wireless & wireline networks 
S. Borst, M. Markakis, Iraj Saniee,
IEEE Trans. on Networking, Vol. 22, Issue: 2, pages: 674-687, April 2014.

Spectral analysis of communication networks using Dirichlet eigenvalues
A. Tsiatas, Iraj Saniee, O. Narayan, Matthew Andrews
Proc. of ACM WWW2013, Rio de Janiero, May 2013.

Bootstrap percolation on random geometric graphs
M. Bradonjic, Iraj Saniee
Extended abstract: Proc. of SIAM Analytic Algos and Combi.Conf, New Orleans, Jan. 2013
Full article: Prob. in Engin. and Inform. Sciences, Cambridge University Press, 2014.

Large-scale curvature of networks
O. Narayan, Iraj Saniee
Physical Review E (statistical physics), Vol. 84, No. 066108, Dec. 2011.

Scaling of load in communication networks
O. Narayan, Iraj Saniee
Physical Review E (statistical physics), Vol. 82, No. 036102, Sep. 2010.

Decentralized control and optimization of networks with QoS-constrained services
Iraj Saniee
Proceedings of the IEEE-ICC 2009, Dresden, June 2009.

Combined network design and multi-period pricing: modeling, solution techniques and computation
D. Bienstock, O. Raskina, Iraj Saniee, Q. Wang
Operations Research, Vol. 54, No. 2, pp261-276, March-April 2006.

Dynamic optimization in cellular networks
S. C. Borst, A. Buvaneswari, L. M. Drabeck, M. J. Flanagan, J. M. Graybeal, G. K. Hampel, M. Haner, W. M. MacDonald, P. A. Polakos, G. Rittenhouse
Iraj Saniee, A. Weiss, P. A. Whiting
Bell Labs Technical Journal, Volume 10, Issue 2
Special Issue on Future Wireless Communications, p 99-119,
Wiley Interscience, Summer 2005.

Load characterization and load anomaly detection for Voice over IP traffic
M. Mandjes, Iraj Saniee, S. Stolyar
IEEE Trans. on Neural Networks, Special Issues on Adaptive Learning Systems in Communication Networks
Edtr. A. Parlos, Vol. 16, No. 5, pp1019-1026, Sep 2005.

Analytic description of shared restoration capacity for mesh networks
M. Bhardwaj, L. McCaughan, S.K. Korotky, Iraj Saniee
Journal of Optical Networking, Vol 4, No. 3, pp130-141, Feb. 2005.

Simplified layering and flexible bandwidth with TWIN
I. Widjaja, Iraj Saniee
Proceedings of Workshop on Future Directions in Network Architecture,
SIGComm 2004, Portland, Oregon, Aug 2004

Collective Behavior: from Cells to Societies, (IDR Team, Group A) Gene Robinson et al, Steering Committee, The National Academies Keck Future Initiative, NA Press, July 2015.

Bell Labs Technical Journal Special Issue on Self-Organizing and Self-Optimizing Networks, Iraj Saniee, U. Barth, Guest Editors, Vol. 15, Issue 3, Dec. 2010.

Complex Systems, (Task Group Summary 9) H. Eugene Stanley et al, Steering Committee, The National Academies Keck Future Initiative, NA Press, 2009.

New Trends in Network Design, Edited by W. Ben-Ameur, D. Bienstock and Iraj Saniee, Special issue of Annals of Operation Research (in conjunction with the International Network Optimization Conference Evry/Paris, 27-29 October, 2003), Vol. 146, No. 1, Sep 2006.

Load balancing in dynamic CDMA networks, S. Borst, G. Hampel, Iraj Saniee, P. Whiting, Handbook of Optimization in Telecommunications, Ed. P.M. Pardalos and M.G.C. Resende, Springer Science, 2006.

Heuristic Approaches for Telecommunications Network Management, Planning and Expansion, Edited by R. Doverspike and Iraj Saniee, Kluwer Academic Publishers, April 2000.

• 11,188,617 Method and network node for internet-of-things (IoT) feature selection for storage and computation
• 10,545,629 Graphical interface for an augmented intelligence system
• 10,142,407 Centralized load balancer with weighted hash function
• 9,104,487 Reducing response time variance of virtual processors
• 9,100,347 Method of burst scheduling in a communication network
• 8,937,957 Intelligent media gateway selection for multimedia communication sessions
• 8,606,274 Method of load-aware dynamic handover in cellular networks
• 8,537,835 Methods and apparatus for self-organized caching in a content delivery network
• 8,055,134Optical telecommunications network and method
• 7,941,156 System of wireless base stations employing shadow prices for power load balancing
• 7,885,803 System and method for simulating traffic loads in packetized communication networks
• 7,729,704 Power load balancing in cellular networks employing shadow prices of base stations
• 7,653,042 Method of burst scheduling in a communication network
• 7,633,865 Network operations control in packet data networks
• 7,573,815 Flow control and congestion management for random scheduling in time-domain wavelength interleaved networks
• 7,561,534 Methods of network routing having improved resistance to faults affecting groups of links subject to common risks
• 7,466,913 Method for WDM optical networks including alternate routes for fault recovery
• 7,340,172 Optical WDM-TDM network
• 7,283,753 System and method for WDM communication with interleaving of optical signals for efficient wavelength utilization
• 7,283,552 Method of scheduling bursts of data for transmission in a communication network
• 7,280,526 Fast and scalable approximation methods for finding minimum cost flows with shared recovery strategies, and system using same
• 7,058,296 Design method for WDM optical networks including alternate routes for fault recovery
• 5,546,542 Method for efficiently determining the direction for routing a set of anticipated demands between selected nodes on a ring communication network