Picture of Rana Ali Amjad

M.Sc. Rana Ali Amjad

Technical University of Munich

Chair of Communications Engineering (Prof. Kramer)

Postal address

Postal:
Theresienstr. 90
80333 München

Research Interests

  • Artificial Intelligence/Machine Learning.
  • Information Theory.
  • Physical Layer Security.
  • Channel and Source Coding Theory.
  • Algorithms and Discrete Mathematics.

Education

  • PhD Student at LNT, TUM since January 2014
  • MSCE 2011-2013, TUM
  • BSc Electrical Engineering and Computer Science, University of Engineering and Technology, Lahore, Pakistan

Awards

  • Walter Gademann Prize for best Master thesis in Department of Electrical, Electronic and Computer Engineering at Technical University of Munich, Germany.
  • Achievement Award for outstanding performance in Master in Communication Engineering degree at Technical University of Munich,  Germany.
  • Naeem Shafi Gold Medal for best performance(Communications Major) in Bachelors of Electrical Engineering and Computer Science degree at University of Engineering and Technology, Lahore Pakistan.
  • Academic Color Holder in High School for best performance over the span of 4 years.
  • Gold Medal for best performance in High School State Examinations.

Theses

Available Theses

Forschungspraxis or MSCE Internships

Code design for Physical Layer Security

Code design for Physical Layer Security

Description

Wiretap channel represents the basic setup for physical layer security. It has been extensively studied in the last four decades and the fundamental limits of communication for this channel are known in a wide variety of scenarios. Nevertheless the only explicit code construction that can achieve wiretap secrecy capacity uses Polar codes. Designing codes for secrecy involve the combined design of codes for reliability and channel resolvability. In 2015 a new coding scheme for channel resolvability was introduced by Amjad and Kramer. The aim of this internship is to combine this channel resolvability code with existing channel codes in order to design wiretap code.

Supervisor:

Theses in Progress

Master's Theses

Designing Clustering Algorithms based on Message Passing Approximation Techniques

Designing Clustering Algorithms based on Message Passing Approximation Techniques

Keywords:
Unsupervised learning, Clustering, Affinity Propagation, Factor graphs, Message passing Algorithms

Short Description:
In this thesis the student will work on designing new low complexity clustering algorithms based on message passing in factor graphs.

Description

Clustering is the task of grouping objects in a ''meaningful'' way such that objects in the same group are more similar to eachother than to those in other groups. Clustering is the main workhorse in un-supervised learning and has numerous applications in various disciplines including data mining and pattern recognition. 

Since it's introduction in 2007, Affinity Propagation (AP) has become one of the most well known clustering algorithms. The main idea of AP is to optimize the final pairwise similarity based clustering cost function in an approximate way by using a low complexity  message passing algorithm. It has been applied to various kinds of datasets ranging from gene expression data to image categorization tasks. Unfortunately AP suffers from the fact that it searches for spherical neighbourhoods to form clusters.

In this work the student will work on designing a new algorithm which in principle is motivated also by AP but is much more effective in discovering non-spherical patterns in data as well. 

Supervisor:

Rana Amjad - Dr. Martin Kleinsteuber (Mercateo/LDV TUM)

Student

Rayyan Khan

Application of Machine Learning in Wireless Communications

Application of Machine Learning in Wireless Communications

Description

This thesis is done in collaboration with Prof. David Gesbert and Dr. Paul de Kerret at Eurecom. 

Machine Learning allows devices to learn from data to improve their strategy for future actions, in either continuous (regression) or discrete (classification) mode. Recently, Machine Learning has evolved into one of the most active research fields due to its impressive success in a wide range of domains where it has often revolutionized the state of the art. Yet, its impact on communication networks has so far been very limited despite a wide array of anticipated application scenarios. 

 
This thesis is part of a project investigating the use of mobile flying base stations, specifically an access point (AP) mounted on Unmanned Aerial Vehicles (UAVs), for wireless communication. Mounting access points on UAVs provides an additional degree of freedom in terms of mobility and system design and offers advantages in terms dynamic network deployment and fast response to varying demand. The goal of this thesis will be to investigate if Machine Learning can be used to solve challenges related to the autonomous positioning of the drone, as the position of the base station is the deciding factor in the system performance. This includes considerations about the Quality of Service, coverage, overall capacity and energy efficiency/flying time, as well as the optimality of static positioning compared to dynamic flight paths. The key features related to the use of Machine Learning to solve this challenge will be studied and the improvement will be quantified whenever possible.

Supervisor:

Rana Amjad - Prof. David Gesbert, Paul de Kerret (Eurecom)

Student

Harald Joachim Bayerlein

Information Theoretic Analysis of Neural Networks

Information Theoretic Analysis of Neural Networks

Description

Various types of neural networks have gained a lot of attention in recent years and have found numerous practical applications with impressive results. Albeit their success, their behaviour is not very well understood mathematically. The aim of this thesis is to approach the topic from an information theoretic perspective and see if one can use insight from information and coding theory to analyze/design neural networks for specific applications.

Supervisor:

Student

Kairen Liu

Forschungspraxis or MSCE Internships

Code Design for Secret Key Generation/ Left over Hash Lemma

Code Design for Secret Key Generation/ Left over Hash Lemma

Description

The source model of secret key generation deals with the idea of Alice and Bob generating a key in a distributed manner from correlated observations. This key must be kept secret from an evesdropper. In this internship/thesis the student will start by looking at a simpler model which corresponds to the left over hash lemma. The student will build on some preliminary work done by me to design codes for the extraction of left over hash in a simple setting. After this the student will (if time permits) extend the work to design codes for simple cases of distributed secret key generation for the source model.

Supervisor:

Student

Amir Hossein Rezaeitabar

Publications

2017

  • Amjad, R. A., Geiger, B. C., Blöchl, C.: Information Theoretic Cost Functions for Markov Aggregation and Clustering. Seminar at Eurecom, France, Oct 2017 more…
  • Geiger, B. C. and Amjad, R. A.: Generalized Kullback-Leibler Aggregation of Markov Chains. Workshop on Information and Communication Theory in Control Systems, May 2017 more…
  • Amjad, R.A.; Geiger, B.: Mutual Information Based Clustering: Hard or Soft? 18th Joint Workshop on Communications and Coding (JWCC), Mar 2017 more…
  • Geiger, B. C.; Amjad, R. A.: Mutual Information-Based Clustering: Hard or Soft? Proc. of 11th ITG Conf. on Systems, Communication and Coding (SCC) (ITG-Fachbericht 268), VDE, Feb 2017, 1-6 more…

2016

  • Amjad, Rana Ali: Variable-to-Fixed Length Resolution Codes for Approximate Random Number Generation. XV International Symposium on Problems of Redundancy in Information and Control Systems, Sep 2016 more…
  • Amjad, Rana Ali: Error Exponent and Achievable Rates for Probabilistic Amplitude Shaping. LNT Tagung, Aug 2016 more…
  • Geiger, B. C. and Amjad, R. A.: Hard Clusters Maximize Mutual Information – Some Results and an Open Problem. Internal LNT Workshop, Aug 2016 more…
  • Amjad, Rana Ali: Information Theoretic Clustering. European School of Information Theory (ESIT), Apr 2016 more…

2015

  • Amjad, R.A.; Kramer, G.: Channel resolvability codes based on concatenation and sparse linear coding. IEEE Int. Symp. Inf. Theory (ISIT), Jun 2015 more…
  • Amjad, Rana Ali: Sparse and Concatenated Codes for Channel Resolvability. Summer School on Information Processing for Large Networks, Jun 2015 more…
  • Amjad, Rana Ali: Low Complexity Codes for Channel Resolvability. Euopean School of Information Theory (ESIT), Apr 2015 more…
  • Amjad, Rana Ali: Coding for Channel Intrinsic Random Extraction. 17th Joint Conference on Communications and Coding (JCCC), Mar 2015 more…
  • Amjad, Rana Ali: Algorithms for Distribution Matching and Resolution Coding. SP Coding and Information School, Jan 2015 more…

2014

  • Amjad, Rana Ali: A Learning Perspective of Context Tree Weighting. Machine Learning in Communication, Sep 2014 more…
  • Böcherer, G.; Amjad, R. A.: Informational Divergence and Entropy Rate on Rooted Trees with Probabilities. IEEE Int. Symp. Inf. Theory (ISIT), Jun 2014 more…
  • Amjad, Rana Ali: Coding Theorems and Algorithms for Simulation of Discrete Memoryless Sources. Euopean School of Information Theory (ESIT), Apr 2014 more…

2013

  • Böcherer, G.; Amjad, R. A.: Fixed-to-Variable Length Resolution Coding for Target Distributions. IEEE Information Theory Workshop (ITW), Sep 2013 more…
  • Amjad, R. A.; Böcherer, G.: Fixed-to-Variable Length Distribution Matching. IEEE Int. Symp. Inf. Theory(ISIT), Jul 2013 more…
  • Bai, Q; Amjad, R. A.; Nossek, J.A.: Average Throughput Maximization for Energy Harvesting Transmitters with Causal Energy Arrival Information. IEEE Wireless Communications and Networking Conference, IEEE, 2013Shanghai, P.R. China, 4232-4237 more…

2012

  • Khalid, Farhan Bin; Amjad, Rana Ali; Chohan, M.A.; Khizar, Muhammad M.: FPGA based real- time signal processor for Pulse Doppler Radar. Conference on Informatics, Electronics and Vision (ICIEV), May 2012 more…