M2 Advanced Wireless Communications Systems

The course is composed of 7 lectures of 3h each. Additional 3h is reserved for solving problems set (TD). Instructor: Prof. Mohamad Assaad Grading: final exam (3h).

This course provides the main concepts and tools required for the design, modelling and optimization of wireless networks. Both Cellular and Local Area Networks will be covered. The focus is on understanding multiuser access techniques and intelligent control of information over the networks. Design and performance analysis of smart scheduling, resource optimization and distributed random access schemes will be addressed in detail in this course. Recent research topics that enable the evolution of wireless systems will be covered in this course as well.

Course outline
1- Introduction to Wireless Networks
Radio Propagation in Mobile Channel, Performance criterion, Duplexing and Multiple Access Techniques
2- Design and challenges of Power control in wireless networks
-Centralized/distributed control for fixed and variable SINR, Industry adoption and Theory limitation, Open problems
3- Scheduling in wireless networks
- Queueing Stability, Throughput optimal scheduling: Max weight policy, Scheduling with delay and fairness constraints, Open problems
3- Resource optimization in multichannel (e.g. multicarrier) networks
- Joint power and channel allocation with instantaneous power and rate constraints, Joint power, user and channel scheduling with long term fairness and throughput constraints,
4- ALOHA and CSMA: performance evaluation and stability
- Performance evaluation of ALOHA, Performance evaluation of CSMA: Bianchi’s model, Stability of ALOHA and CSMA, Rate optimal CSMA.

Queueing Theory, Markov Chains, Optimisation.

?S. Stanczak, M. Wiczanowski and H. Boche, Fundamentals of Resource Allocation in Wireless Networks, Second edition, Springer 2009. ?Michael Neely, Stochastic Network Optimization with Application to Communication and Queueing Systems, Morgan & Claypool Publishers ?Vincent K. N. Lau and Yu Kwong Ricky Kwok, Channel-Adaptive Technologies and Cross-Layer Designs for Wireless Systems with Multiple Antennas, Wiley, 2006, ?Gast, 802.11 Wireless Networks: The Definitive Guide, O’Reilly ?Kelly, Reversibility and Stochastic Networks, Cambridge Mathematical Library ?Bianchi, Performance analysis

Octobre - Novembre - Décembre.

GIF-SUR-YVETTE

The course is composed of 6 lectures of 3h each and two practical work of 3h each. Grading: final exam (3h).

The goal of this course is to provide students with an introduction to the field of machine learning, with a particular focus on online learning. In the first part we will review statistical learning theory, and provide efficient algorithms for regression and classification. In the second part, we will focus on reinforcement learning which enables to solve Markov Decision Processes (MDPs) when the statistics of the environment are unknown. We will first consider algorithms for Multi-Armed Bandits (i.e. MDPs with a single state), and then extend our approach to general MDPs. Approximation techniques for large scale problems will also be addressed. Applications in the context of wireless networks will be discussed.
Tentative Syllabus
I Statistical Learning
– Framework, empirical risk minimisation, overfitting
– Algorithms: gradient descent, stochastic gradient descent
– Statisticalmodels:linearclassification/regression,kernelmethods,neuralnetworks
II Reinforcement Learning I (
– Multi-Armed Bandits (MAB)
– Algorithms for stochastic MABs: UCB, KL-UCB, Thompson Sampling – Algorithms for adversarial MABs: Hedge, EXP3.
III Reinforcement learning II
– Basic algorithms: Q-Learning, SARSA
– Algorithms for large scale problems: value approximation, policy gradient. – Monte-Carlo tree search and the UCT algorithm.
IV Application of Learning in Wireless Networks
– Example of wireless networking problems where Reinforcement Learning is ap-
plied.
– Examples of Implementation using Matlab, Python.

Probability, statistics, optimisation.

1. Machine Learning: A Probabilistic Perspective, by Kevin P. Murphy, Adaptive Computation and Machine Learning series, MIT Press, ISBN-10: 0262018020. 2. Machine Learning, by Tom M. Mitchell, Mcgraw-Hill Series in Computer Science, McGraw-Hill Education, ISBN-10: 0070428077. 3. Recent papers in the domain.

Décembre - Janvier.

GIF-SUR-YVETTE

8 units. A single unit ?addresses a specific TOPIC ?does not necessarily correspond to a single lecture ?6h of practical exercices for network usage and modeling ! Lecture slides (mostly before lecture) Reading, supplementary literature Assignments! Grading.

Give to students the basics of networking techniques, their models and their performance. This course focuses on the fundamentals of the wired and wireless communication networks. The focus is on "how things work" and on the "design principles". With the understanding of these basics, the operation of the Internet and the key networking technologies/components (Internet protocols, Congestion Control, Application, Video and streaming, future Internet, etc.) will be discussed. The course covers both the architectural principles for making these networks scalable and robust, as well as the key techniques essential for analyzing and designing them. Thus, performance can be studied using graph theory, Markov chains, queuing, etc.
DETAILED PROGRAM
Background of wireless and mobile communication networks. Physical and MAC Standards.
Introduction to Networks
?Packet Switching Models
?Architecture, layers, protocols
Fundamentals of Internet
?Networking : IPv4 and v6 protocols, Addressing, Relaying, Routing
?TCP : Reliability, Congestion Control
?Applications and Services
Multimedia
?Multimedia and streaming applications
?Multicast
?Quality of Service in Internet
Mobile Networking
?Mobile IP and Mobility Management Protocols
?TCP over Wireless and Mobile Networks
?Cross layer approaches
Future Internet
?Naming, addressing
?Software Defined Networking
?Network Function Virtualization.

Good skill in Communication Networking and TCP/IP protocols, wireless and mobile communication networks : physical techniques, MAC standards, cellular standards.

?J. F. Kurose and K. W. Ross, Computer Networking, A Top-Down Approach Featuring the Internet, Addison Welsey, Last Ed. ?A. Tanenbaum and J. Wetherall, Computer Networks, 5/E , Prentice Hall, 2011 ?W. R. Stevens, TCP/IP Illustrated, protocols - Volume 1, Addison Wesley, ?A.Leon-Garcia & I.Widjaja, Communications Networks, 2nd Edition, McGraw-Hill ?John DAY, “Patterns in Network Architecture. A Return to Fundamentals”, Pearson Education, 2008 ?[She95] Shenker, S. "Fundamental Design Issues for the Future Internet." IEEE Journal on Selected Areas in Communications 13, no. 7 (September 1995)

Novembre - Décembre.

GIF-SUR-YVETTE

The course is composed of 7 lectures (3h each) and 3h for solving problems sets (TD). Grading: final exam (3h).

This is a graduate-level introduction to the fundamental ideas and results of information theory. The course moves quickly but does not assume prior study in information theory. It is intended for graduate students from mathematics, engineering or related areas wanting a good background in fundamental and applicable information theory. Roughly the first third of the course discusses elementary measures and properties of information at a more sophisticated level. The middle third discusses method of Types and the main results of Shannon’s theory, manly the coding theorems for source and channel coding. The remainder touches on topics that are explored more fully in later courses, converses, capacity of wireless channels, etc.
COURSE OUTLINE
1. Properties of Shannon's Information Measures
2. Markov Chain, Fundamental Inequalities and Entropy of Stationary Sources
3. Lossless Source Coding
Weak Typical Sequences, Noiseless Source Coding and its Coding Theorem,
4. Method of Types
Type Counting Lemma, Strong Typical Sets, Delta Sequences, etc.
5. Coding Theorem for Noisy Channels (Binary Symmetric Channels, Gaussian Channels (AWGN), etc.)
6. Coding Theorem for Lossy Source Coding
7. Converse to the Coding Theorems for Discrete Memoryless Sources (DMSs) and Channels (DMCs)
8. Capacity of Wireless Channels
Channels with Colored Gaussian Noise and Fading, Channels with Multiple Antennas (MIMO) and Fading, etc.

Analysis, linear algebra and probability at undergraduate level.

[1] COVER, T.M., and THOMA S, J.A., Elements of Information Theory, Wiley, 1991. [2] RIOUL, O., Théorie de l’Information et du Codage, Hermes Science – Lavoisier, 2007. [3] CSISZAR, I., and KORNER, J., Information Theory: Coding Theorems for Discrete Memoryless Systems, Academic Press, 1997. [4] ASH, R.B., Information Theory, Interscience Publishers, 1966. [5] GOLDSMITH, A., Wireless Communications, Cambridge University Press, 2005. [6] TSE, D., VISWANATH, P., Fundamentals of Wireless Communications, Cambridge University Press, 2005. [7] GALLAGER, R.G., Information and Reliable

Octobre - Novembre - Décembre.

GIF-SUR-YVETTE

6 lectures (18H) -Part 1 : point-to-point channel oCapacity and outage of wireless channels oDiversity and multiplexing, space-time codes oReceiver architecture : maximum likelihood detection, linear detection, successive interference cancellation (SIC) -Part 2 : multi-user network oMultiple access channel (uplink) : capacity region, SIC oBroadcast channel (downlink) : capacity region, dirty paper coding, beamforming oOther networks : interference channels, relay channels, network MIMO, massive MIMO.

Practical Work (6H) : -EL1 : Channel capacity, waterfilling power allocation -EL2 : Space-time transmission, performance of different receivers -EL3 : Random beamforming and scheduling in broadcast channels.

This course introduces the information theory for multi antennas systems in wireless networks. The objective of this course is to study the fundamental limits of multi antennas communications in a multi-user environment with fading and noise, as well as the corresponding communication strategies.

Outline
Part 1 : point-to-point channel
oCapacity and outage of wireless channels
oDiversity and multiplexing, space-time codes
oReceiver architecture : maximum likelihood detection, linear detection, successive interference cancellation (SIC)
-Part 2 : multi-user network
oMultiple access channel (uplink) : capacity region, SIC
oBroadcast channel (downlink) : capacity region, dirty paper coding, beamforming
oOther networks : interference channels, relay channels, network MIMO, massive MIMO.

Digital communications, Information theory, Linear Algebra.

-Tse and Viswanath : Fundamentals of wireless communications -EL Gamal and Kim : network Information Theory.

Décembre - Janvier.

GIF-SUR-YVETTE

Jocelyn Fiorina, Mohamad Assaad, Sheng Yang.

The course is taught by J. Fiorina, M. Assaad and S. Yang. It is composed of 7 lectures (3h each) and 3h for solving exercices (TD).

This course aims at providing students a sound knowledge of modulation theory and reception theory. Emphasis will be placed on specific signalling concepts for single-user transmission over wireless channels.

Course Outline

- Deterministic signals: Signals as vectors, geometry of linear spaces, expansions and transforms, modulation, demodulation, bandpass signals, sampling, reconstruction, ISI, Nyquist criterion, matched filter, performance analysis.
- Digital modulations: Linear modulations, constant envelop modulations, mapping.
- Detection and equalization in single-carrier systems: Maximum Likelihood Sequence Estimation for linear AWGN channels. Sub-optimal equalization: Linear equalization (ZF, MMSE), Decision-Feedback Equalization (DFE), Decision-Feedback Sequence Estimation (DFSE).
- Digital communications through fading (multipath) channels: The concept of diversity (frequency, time and spatial diversity), performance analysis.
- Spread spectrum communications: Direct sequence, frequency-hopping.
Multichannel and multicarrier systems, OFDM.
- Ultra Wide Band (UWB) techniques, impulse radio.

Course "Mathematical Basis for Communications" or equivalent Course "Information theory" or equivalent.

[1] BENEDETTO, S., and BIG LIERI, E., Principles of Digital Transmission with Wireless Applications, Kluwer Academic Plenum Publishers, 1999 [2] PROAKIS, J., Digital Communications, McGraw-Hill, 2000 [3] HAYKIN, S., Communication Systems, Wiley, 2002 [4] TSE, D., VISWANATH, P., Fundamentals of Wireless Communications, Cambridge University Press, 2005.

Septembre - Octobre.

GIF-SUR-YVETTE

7 Lectures of 3h each. One 3h is reserved for solving Problem Sets (TD).

This course will be tool oriented. The objective is to introduce stochastic models and algorithms through their application, mainly in the channel estima- tion area. Important tools in linear and non linear filtering, classification and learning are derived and applied.
Syllabus:
1/Propagation in random media, antennas, models of signal reception. 2/Introduction to MIMO link design: information theoretic principles. 3/Discrete and continuous-state Markov models and their estimation: Dynamic programming algorithms, Viterbi and forward-backward proce-
dures.
Extension to Markov graphical models, spatial models. Cluster-variation
methods. Pseudo-likelihood estimation.
EM and general variational procedures. Baum-Welsh algorithm.
Kalman filtering. Nonlinear evolution and filtering in discrete time: the
change of measure method.
Continuous evolution-discrete observation models. Basics of stochastic inte-
gration and the evolution of the conditional expectation.
Particle filters and general dynamic Monte-Carlo methods.
4/ Applications to blind and semi-blind equalization, rapidly time-varying
channels.
5/ Fundamentals of classification methods. Application to code and channel
detection.
6/ Introduction to learning algorithms. Self-learning and control of Markov
chains.

Probability, Markov Chains.

Recent papers in the domain.

Octobre - Novembre - Décembre.

GIF-SUR-YVETTE

Antoine Berthet, Frédéric Dufaux.

The course is composed of two parts: channel coding (30h) and source coding (6h) Lecture 1-Introduction to error-correcting codes (3h) TD1:error-correcting codes (1.5h) Lecture 2-Linear cyclic codes (3h) TD2: Algebraic decoding (Euclid or Berlekamp Massey) (1.5h) Lecture 3-Linear convolutional codes (3h) Lecture 4-Performance of linear codes under MLD (3h) Lecture 5-Factor graphs and the sum-product algorithm (3h) Lecture 6-Sparse-graph codes (3h) TD3: Programming exercise on LDPC codes (1.5h) 7-Sparse-graph codes (3h) - Density evolution, Code design optimization under iterative decoding TD4: Programming exercise on turbo codes (1.5h) 8- Sparse-graph codes (3h) -Ensemble enumerators, Code design optimization under MLD Source coding (4.5h+1.5h TD5) 1. Variable-length lossless compression, Fano-Shannon coding 2. Transform coding 3. Predictive coding 4. Introduction to video compression TD5: source coding (1.5h).

The field of channel coding started with Claude Shannon’s 1948 landmark paper. Almost seventy years of efforts and invention have finally produced coding schemes that closely approach Shannon’s channel capacity limit on AWGN channels, both power-limited and band-limited. Similar progress is being achieved in other important applications such as single-user and multiuser wireless channels. In the first part of the course, we remind students of the basics of the algebraic coding theory for conventional binary-input output-symmetric memoryless channels. The second part of the course is devoted to sparse graph codes. We review in detail code construction aspects, iterative decoding, and mathematical tools for design optimization. We then expound the principles of non-binary coding for the Gaussian channel and show how and why coded modulations can also benefit from sparse-graph codes optimized for binary-input channels and iterative decoding.

[1] GALLAGER, R.G., Information Theory and Reliable Communications, Wiley, 1968. [2] COVER, T.M., THOMAS, J.A., Elements of Information Theory, John Wiley, 1991. [3] BERLEKAMP, E., Algebraic Decoding Theory, McGraw Hill, 1968. [4] MAC WILLIAMS, F.J., SLOANE, N.J.A., Theory of Error-Correcting Codes, North Holland Publishing, 1977. [5] MCELIECE, R.J., Finite Fields for Computer Scientists and Engineers, Kluwer Academic Publishers, 1987. [6] VITERBI, A.J., OMURA, J.K., Principles of Digital Communications and Coding, McGraw Hill, 1979. [7] SCHLEGEL, C.B., PEREZ, L.C., Trellis and Turbo Coding

GIF-SUR-YVETTE

Jean-Pierre Barbot, Thomas Rodet.

The course is divided into two parts of 12h each. The first part, taught by T. Rodet, focuses on Probability Theory while the second one is taught on J.-P. Barbot and focuses on estimation theory. Each composed of 3 lectures of 3h each and additional 3h reserved for exercices (TD).

The aim of the first part of this course is to refresh basic notions on probability theory, such as independence and conditional probabilities, probabilities on finite and countable space (random variables, moments, classical distributions, Bernoulli, Binomial, Poisson, etc), random variables on R (Gaussian, Exponential, etc), independent random variables and characteristic functions, the multivariate Normal distributions and different types of convergence (Central Limit Theorem and the Laws of Large Numbers). Many examples on signal processing and/or telecommunications and/or game theory will be given to illustrate the probability theory. The second part aims at providing students with a sound knowledge in estimation (decision) theory.
COURSE OUTLINE
- Introduction to probabilities on events.
- Probabilities on a finite or countable space
- Random variables on continuous space (R). Independent random variables and characteristic functions.
- Multivariate distributions (Gaussian distribution).
- Different convergences (Central Limit Theorem and the Laws of Large Number).
- Introduction to parameter estimation and confidence areas
- Concept of sufficient statistics
- UMVU (Uniformly Minimum-Variance Unbiased) estimators
- Cramér-Rao bound, Fisher information
- Maximum Likelihood estimator, asymptotic efficiency
- Bayesian estimators (MMSE, L1 Norm, MAP)
- Posterior Cramér-Rao bound
- Detection: Neyman-Pearson lemma
- Bayesian detection.

Analysis, linear algebra and probability at undergraduate level.

[1] JACOD, J. AND PROTTER, P.E., Probability Essentials, Springer Verlag, 2003 [2] BREMAUD, P., An Introduction to Probabilistic Modeling, Springer, 1988 [3] CHUNG, K.L., A Course in Probability Theory, Academic Press, 2001 [4] FELLER, W., An Introduction to Probability Theory and its Applications, Vol I & II - John Wiley & Sons, Wiley Series in Probabilities and Statistics, 1968 [5] LEHMANN E.L., and CASELLA, G., Theory of Point Estimation, Springer Texts in Statistics, 2003 [6] LEHMANN, L., and ROMANO, J.P., Testing Statistical Hypotheses, Springer Texts in Statistics, 2006 [7] VAN

Septembre - Octobre.

GIF-SUR-YVETTE

Samson Lasaulce, Koen de Turck.

The course is divided into two parts of 12h each: Convex optimization and Queueing Theory. The convex optimization part is taught by S. Lasaulce while Queueing Theory is taught by K. de Turck. Eight 3-h classes over the period Septembre-October. Grading: final exam.

The first purpose of this class is to introduce fundamental tools which are used in typical optimization problems encountered when analyzing and designing communication networks. Then, an introduction of queuing theory will be given. The main part of the
courses consists in an overview of the mathematical tools used in queuing theory, in
particular Markov chains and Markov processes. The applications of queuing theory to
telecommunications network will be illustrated through practical exercises.

Course outline
Review of fundamental notions of set theory and topology
Unconstrained optimization
Constrained optimization
KKT conditions
The important case of convex optimization, global optimality, duality principle
A short note on optimization in dynamical systems: the dynamic programming principle (Viterbi algorithm example), HJB equation
Numerical methods for optimum design
Distributed optimization in communication networks
Basic notions of games and learning
Markov Chains
Markov Processes Little’s Formula.
M/M/1, M/M/m, M/M/m/m, M/M/1/K, M/M/1/-/K, M/M/-/-/K Queues M/GI/1 and GI/M/1
Jackson network
Application to communication networks
Simulation of queues (Workshop).

Mathematical background of BSc students in electrical engineering.

[1] ARORA, J. A, Introduction to Optimum Design, Elsevier Academic Press, 2nd edition, 2004. [2] BOYD, S., and VANDENBERGHE, L., Convex Optimization, Cambridge University Press, 2004. [3] HIRIART-URRUTY, J.-B., and LEMARECHAL C., Convex Analysis and Minimization Algorithms I and II, Springer-Verlag, 1993. [4] HADJISAVVAS, N. et al, Handbook of Generalized Convexity and Generalized Monotonicity, Springer, 2005. [5] ENGWERDA, J. C., LQ Dynamic Optimization and Differential Games, Wiley, 2005. [6] An Introduction to Queuing Theory: and Matrix-Analytic Methods - by L. Breuer and Dieter Baum

Septembre - Octobre.

GIF-SUR-YVETTE

- The internship is a Full-time work starting at the beginning of April. The minimum duration is 4 months, and the typical duration is 5 months.
-For students from partners schools and universities in France and abroad, the internship duration is fixed s.

Aout - Septembre - Avril - Mai - Juin - Juillet.

All instructors involved in SAR program.

-Topics are typically provided by the academic staff involved in the master program, but the student can also work on a project provided by partner R&D labs. -Each project covers a research topic currently under investigation by the instructor. -Two stud.

The students will work on research projects proposed by the academic staff involved in the master program. The topics of the projects are related to the research activities of the instructors. The topics can cover different aspects of wireless communications and networking. The students have two months to understand the project topic and perform the required tasks.

Wireless communications, networking, etc.

Recent papers in the domain.

Février - Mars.

GIF-SUR-YVETTE

All instructors involved in the master program, L2S members, invited researchers from academia and industry.

Each year new seminars will be proposed to the students. In total 4 seminars (of 3h) each will be presented each year.

Grading: Each student has to write a summary of typically 2-3 pages about the seminars.

The objective is to provide the students with detailed presentations of specific research topics. The topics covered by the seminars are up-to-date subjects currently under studies and development by the presenters. The objective is to help students understanding current research activities in the domains of wireless communications and networking, 5G and 6G systems, applications of machine learning in wireless systems, as well as other research topics.

Background in Wireless communications and networking.

Recent papers in the domain.

Février - Mars.

GIF-SUR-YVETTE

This course will cover the following subjects in four lectures (3h each): 1. Introduction to Neural Networks: - Perceptrons, capacity of a single neuron, Linear and logistic regression, Backpropagation and stochastic gradient optimisation 2. Advance Neural Networks: -Auto-encoders and variants, hyper-parameters and training tricks for neural networks, Unsupervised learning of representations and pre-training 3. Variations on auto-encoders and probabilistic Graphical Models: -Convolutional neural networks, Recurrent networks, Variational auto-encoders and Generative Adversarial Nets, Review of directed models (HMMs and mixtures), Inference, sampling and learning algorithms for such models, Deep belief networks, deep Boltzmann machines 4. Modern architectural variations for communications and IoT data analytics: - A quick view of Theano, Typical classification tasks: classifying MNIST digits using logistic regression, Deep learning for solving communication problems.

This is a course on the principles of representation learning in general and deep learning in particular. Deep learning has recently been responsible for a large number of impressive empirical gains across a wide array of applications including most dramatically in object recognition and detection in images, natural language processing and speech recognition. On the other hand, Internet of thing (IoT) and Smart City deployments are generating large amounts of time-series sensor data in need of analysis. Developing deep learning methods to these domains is an important and very timely research topic.
In this course we will explore both the fundamentals advances in the area of deep learning and the recent applications to the field of IoT and in general communications. Our focus will be on recent applications of deep learning to perform data analytics on Internet of Things (IoT) communications, including neural networks, auto-encoders, convolutional neural networks and recurrent networks. We willalso consider well-known probabilistic graphical models, including undirected models and directed models that have recently shown promise (e.g. Boltzmann machines, Deep Believe Nets).
This course will cover the following subjects in four lectures:
1. Introduction to Neural Networks:
2. Advance Neural Networks:
3. Variations on auto-encoders and probabilistic Graphical Models: 3.1. Convolutional neural networks
4. Modern architectural variations for communications and IoT data analytics: 4.1. A quick view of Theano.

Information theory, probability.

1. Deep Learning, by Ian Goodfellow, Yoshua Bengio and Aaron Courville, Adaptive Computation and Machine Learning series, MIT Press, November 2016, ISBN-10: 0262035618. 2. Machine Learning: A Probabilistic Perspective, by Kevin P. Murphy, Adaptive Computation and Machine Learning series, MIT Press, ISBN-10: 0262018020. 3. Machine Learning, by Tom M. Mitchell, Mcgraw-Hill Series in Computer Science, McGraw-Hill Education, ISBN-10: 0070428077. 4. Deep Learning - Methods and Applications by Li Deng and Dong Yu http://research.microsoft.com/pubs/219984/BOOK2014.pdf.

Février - Mars.

GIF-SUR-YVETTE

The course is taught by S. Lasaulce. Four 3-h classes over the period February-March. The exam consists in summarizing and presenting a research paper.

One of the goals is to provide an overview of game theory, in particular of : direct game theory and mechanism design ; the main mathematical models (normal form, extensive form, coalitional game) ; the main solution concepts ; the methodologies for equilibrium analysis ; learning in games. Most of the notions are illustrated by wireless examples and several wireless case studies are analyzed in detail.

Basic notions of optimization (B.Sc level).

S. Lasaulce and H. Tembine, « Game Theory and Learning for Wireless Networks : Fundamental and Applications », Academic Press.

Février - Mars.

GIF-SUR-YVETTE

The course is composed of 4 lectures (3h each).

Organization of the seminar: 1 Introduction 1.1 What is IoT ? 1.2 Applications 2 Design requirements and challenges 3 MAC and Routing refreshers 4 Wireless Sensor Networks 4.1 MAC 4.2 Routing 4.3 A quick look at the standards 5 LPWANs 6 Cellular IoT 7 Performance evaluation 7.1 Modeling energy consumption 7.2 Stochastic models 7.3 Discrete event simulation 7.4 Experimentation

Grading: A research paper to be summarized and presented by the students.

In this seminar, we look at the concept of Internet of Things (IoT). We are interested in the networking aspects of IoT: how to enable billions of things to communicate?
We first take time to review the different definitions of the IoT and expose the issues that appear at the link and routing layers. Indeed, several problems are arising when designing network protocols for highly constrained wireless networks such as IoT networks. We then focus on actual network architectures and protocols. We will notably be interested in WSN for which an abundant literature exists. Additionally, we also introduce cellular networks and LPWANs for object connections.
We will discuss the many issues that arise when designing network protocols for a large number of low capacity objects when:
* a limited amount of energy is available
* nodes are equipped with a slow CPU clock and low memory
* they dispose of a low power unreliable radio
* it is not possible to manually (re)configure the nodes
We will be interested in the issues these constraints imply at the MAC and routing layers and review some of the solutions proposed in the scientific literature to address these constraints. Some questions we will explore are: how to provide energy efficient scalable medium access ? How to build and maintain a logical network topology (notably for convergecast traffic) which scales up and do not drain the nodes battery while being self- organized ? What are the tools available to evaluate the performances of such protocols ?.

Background on Internet protocols and Networks architectures.

1. Bachir, A., Dohler, M., Watteyne, T., and Leung, K. K. (2010). Mac essentials for wireless sensor networks. Communications Surveys & Tutorials, IEEE, 12(2) :222– 248. 2. Doherty, L. and Pister, K. S. J. (2008). Tsmp : Time synchronized mesh protocol. IEEE/IFIP DSN, pages 391–398, Orlando, USA. 3. Liu, Chi Harold, Bo Yang, and Tiancheng Liu. "Efficient naming, addressing and profile services in Internet-of-Things sensory environments." Ad Hoc Networks 18 (2014): 85-101. 4. IEEE (2011). Ieee standard for local and metropolitan area networks - part 15.4 : Low-rate wireless personal area

Février - Mars.

GIF-SUR-YVETTE

Michel Kieffer, Pierre Duhamel.

-The course is composed of 4 lectures of 3h each -Grading: Research papers to be summarized and presented by the students.

Traditionally, cross-layer and joint source-channel coding were seen as incompatible with classically structured networks but recent advances in theory changed this situation. Joint source-channel decoding is now seen as a viable alternative to separate decoding of source and channel codes, if the protocol layers are taken into account. A joint source/protocol/channel approach is thus addressed in this seminar: all levels of the protocol stack are considered, showing how the information in each layer influences the others.
This seminar presents the tools to show how cross-layer and joint source-channel coding and decoding are now compatible with present-day mobile and wireless networks, with a particular application to the key area of video transmission to mobiles. Typical applications are broadcasting, or point-to-point delivery of multimedia contents, which are very timely in the context of the current development of mobile services such as audio (MPEG4 AAC) or video (H263, H264) transmission using recent wireless transmission standards (DVH-H, DVB-SH, WiMAX, LTE).
Tentative syllabus
• Introduction: Principles of a reliable reception
• Identifying redundancy
• Structuring redundancy
• Exploiting the redundancy
• Applications
• Extensions, joint-source-channel coding
• Conclusions and open issues in joint protocol-channel decoding.

Information theory, coding, probability.

[1] DUHAMEL, P., and KIEFFER, M., Joint source-channel decoding: A cross-layer perspective with applications in video broadcasting. Academic Press, 2009 [2] SAYOOD, K., Introduction to Data Compression, Academic Press, 2000 [3] COVER, T., and THOMAS, J., Elements of Information Theory, Wiley, 1991.

Février - Mars.

GIF-SUR-YVETTE

Mohamad Assaad, Salah Eddine Elayoubi.

The course is taught by M. Assaad and S. Elayoubi; It is composed of four lectures of (3h) each. The lectures cover the following topics: Lecture 1 I General introduction: wireless network and optimization Lectures 2 and 3: Markov Decision Processes (Definition and objectives, Solution of Bellman equation: Value / Policy iteration, etc.) Lecture 4: Lyapunov based optimization Grading: The exam consists in summarizing and presenting a research paper.

The goal of this course is to give an introduction on how network optimization tools can be applied to solve problems in present and future wireless networks. The first part of this course provides a general description of some problems in wireless networks (mainly in 5G systems and beyond) that require the use of network optimization. It will be shown how the design of the network can be optimized by using a stochastic modeling of the system. In the second part, Markov Decision Processes (MDPs) will be investigated, and efficient algorithms to solve MDPs when the statistics of the environment are known will be given. An alternative approach based on Lyapunov optimization will be provided as well.
Tentative Syllabus
I General introduction
– Wireless networks and optimization, 5G: Physical layer (frame, massive mimo, MTC, etc.), Optimization of the system: stochastic modeling and formulation of some optimization problems
II Markov Decision Processes I
– Definition and objectives, Optimal policy, Bellman equation, complexity
III Markov Decision Processes II
– Solution of Bellman equation: Value / Policy iteration, solutions with low complexity: e.g. weak interference, etc.
IV Lyapunov based optimization.

Optimization, queueing theory, Markov chains.

1. Michael Neely, Stochastic Network Optimization with Application to Communication and Queueing Systems, Morgan & Claypool Publishers 2. Markov Decision Processes: Discrete Stochastic Dynamic Programming. Author( s):. Martin L. Puterman. First published:15 April 1994.

Février - Mars.

GIF-SUR-YVETTE

-The course is composed of 4 lectures of 3h each. -Grading: Research Papers to be summarized and presented by the students.

In this seminar, the basics of random matrix theory will be layed out, and the connection between random matrices and wireless communications will be stressed. The seminar will be composed of two parts: a theoretical introduction to random matrices in the one hand, and the application of some important results to the field of wireless communications in the other. The mathematical prerequisites for this seminar are a good understanding of the various convergence types in probability theory as well as some basics in linear algebra. Only elementary knowledge of multi-user MIMO systems will be required for the application part.

Seminar outline
Introduction to random matrix theory
? Historical review and motivation
? The Marcenko-Pastur law
? The Stieltjes transform approach
? Some important results
Application to wireless communications
? Analysis and optimization of CDMA systems
? Point-to-point MIMO communications
? Multi-user MIMO communications in the uplink and downlink.

Probability, Linear Algebra, Digital communications, MIMO.

[1] R. Couillet, M. Debbah, “Random matrix theory methods for wireless communications”, Cambridge University Press, 2011 [2] Z. Bai, J. W. Silverstein, “Spectral Analysis of Large Dimensional Random Matrices”, Second Edition, Springer, New York, 2009.

Février - Mars.

GIF-SUR-YVETTE

- The course is divided into two parts: Stochastic Geometry Theory and Applications in Wireless Networks -The course is composed of 4 lectures (3h each) -Grading: Research papers to be summarized and presented by the students.

The proposed course provides a rigorous introduction to stochastic geometry, which will enable the attendees to obtain powerful, general estimates and bounds of wireless network performance and make good design choices for future wireless architectures and protocols that efficiently manage interference effects. Several practical engineering applications are discussed and integrated with mathematical theory. In particular, the attendees will learn how to model distributed wireless systems using spatial point processes and their properties. From the basic spatial model, specific aspects of wireless communications such as interference, outage and throughput will be discussed, showing the most common methods to assess their system-level performance. This course provides the necessary background to all academic researchers and practitioners who are interested in the modeling, analysis, and design of future wireless networks. Practical examples using Matlab and R will be illustrated with interactive & hands-on sessions.

Program Topics: Part I - Theory
• Introduction to spatial point processes
• Marked processes
• Poisson and non-Poisson point processes
• Random tessellation
• Palm theory
• Characterizing networks with fading channels
• Methodologies for system-level performance evaluation and optimization
Program Topics: Part II - Applications
• Ad hoc networks
• Cellular networks
• Heterogeneous networks (small cells)
• mmWave networks
• Validation with experimental data.

Basic knowledge of wireless communications, probability, Linear Algebra.

Recent Papers in the domain.

Février - Mars.

GIF-SUR-YVETTE

-4 lectures (3h each) -Grading: Research papers to summarize and present.

Vehicular networking and Vehicular Ad Hoc Networks (VANETs) have emerged as an exciting research and application area. The target applications envisioned for Intelligent Transportation Services concern major societal problems like public safety, traffic management and coordination, air pollution emission reduction or law enforcement but also classical broadband services to users.
Through these objectives some inherent VANET characteristics such as highly dynamic topology, frequently disconnected network, and different and dynamic network density, make communications and data transmission a challenging task in these networks. In this seminar, we provide a comprehensive and critical review of proposed applications, problems and technical solutions for vehicular networking and identify the main challenges in this area.
Course outline
? Applications for VANETs
? Vehicular communications
? Communication Technologies
? VANET
? Routing in VANET
? Location Service
? Mobility Models
? Clustering
? Standardization.

Background on Internet protocols and Networks architectures.

K. Zheng, Q. Zheng, P. Chatzimisios, W. Xiang, and Y. Zhou, \Heterogeneous Vehicular Networking: A Survey on Architecture, Challenges, and Solutions," IEEE Communications Surveys Tutorials, vol. 17, pp. 2377{2396, Fourthquarter 2015. C. Cooper, D. Franklin, M. Ros, F. Safaei, and M. Abolhasan, “A Comparative Survey of VANET Clustering Techniques,” IEEE Communications Surveys Tutorials, vol. PP, no. 99, pp. 1–1, 2016. “Intelligent Transport Systems (ITS);Vehicular Communications; Basic Set of Applications; Part 2: Specification of Cooperative Awareness Basic Service," ETSI EN 302 637-2 V1.3.1

Février - Mars.

GIF-SUR-YVETTE