Research Spotlight

Spotlight: We have developed techniques for the design of static multi-hop wireless networks for sensor interconnection, with application to the design and deployment of the Internet of Things.

Faculty Name: Anurag Kumar


Spotlight: As-you-go deployment of wireless relay networks for sensor interconnection: development of algorithms for one and two connected design, and their experimental evaluation. We have developed algorithms for as-you-go deployment on a line, by modeling the problem as a constrained Markov decision processes. Learning algorithms have been developed, that do not need parameters of the RF propagation environment, by using two time-scale stochastic approximation.

Faculty Name: Anurag Kumar


Spotlight: Spectrum sensing for cognitive radio. [Joint work with Sanjeev Gurugopinath] We considered the case where the primary transmitter’s signal occupies a large bandwidth, but with a high power narrowband component. Such a signal opens up two possibilities for detection: one could either try to detect the entire wideband signal, or one could filter out only the narrowband component and detect its presence or absence. The fact that the signal propagation characteristics of narrowband and wideband signals are quite different leads to an interesting tradeoff between the two options. We analytically characterized this tradeoff, and determined which of the two options offers better detection performance. Achievable rates and outer bounds for the interference channel with cooperation and privacy constraints. [Joint work with Parthajit Mohapatra] In this line of research, we first developed algorithms for interference alignment in a multiuser MIMO interference constant channel setup. Next, we analyzed the generalized degrees of freedom of the K-user MIMO interference channel. Finally, we considered the 2-user interference channel with rate-limited cooperation between the transmitters and secrecy constraints at the receivers. We were able to derive achievable schemes and outer bounds on the rate pairs first for the so-called deterministic approximation of the interference channel, and then extended the results to the Gaussian interference channel. Algorithms for sparse signal recovery. [Joint work with Ranjitha Prasad] We developed algorithms for wideband channel estimation exploiting the sparsity of the channel in the time-domain. Starting from algorithms for channel estimation in the single-antenna case, we extended our algorithms to track time-variations in the channel, to exploit “cluster sparsity”, where the signals arrive in clusters of closely separated delays, and to multiple antenna systems where the signals between different pairs of transmit and receive antennas have a common sparsity structure (support). This latter work won the best paper award in the National Communications Conference 2014, which was held at IIT Kanpur.

Faculty Name: Chandra R.Murthy


Spotlight: In 2014 we have conducted research and development work in different aspects of Computational Electromagnetics: Faster Design Tools: (a) Learning based acceleration in design iterations: In a typical design cycle many iterations on the package-board-system layout may be performed to meet design specifications. In the process, the analysis step needs to be repeated as many times as the number of layout variants. The cost of analysis, especially if using a 3D fullwave extraction methodology, therefore becomes prohibitive for large-scale analysis in the design process. We propose a methodology to expedite analysis of subsequent layout iterations based on information stored from previous layout solution. In this work, eigen-vectors from a previously solved layout is augmented to the Krylov subspace to expedite the convergence of a Generalized Conjugate Residual (GCR)-based iterative solution. Numerical results demonstrate up to 40% improvement in the convergence properties using the proposed GCR method. We presented our findings in 2 international conferences. (b) Neural networks in time-domain design: Signal speeds of high speed serial data links double almost every generation and with increasing speeds, simulation and modeling challenges are getting more complex. The present popular and widely accepted metric for simulating a high speed link from signal integrity (SI) perspective is Bit Error Rate (BER) testing. End to end channel simulation in frequency domain is much faster and less resource intensive. We propose a multi-layer perceptron (MLP) type of artificial neural network (ANN)-based mapping to predict eye-height from frequency domain metrics like return loss (RL) and insertion loss (IL). We presented our findings in 2 conferences and our paper won best student paper award at IEEE EDAPS 2014. Faster Simulation Engines: (a) Hardware acceleration of EM solver: We propose a Field Programmable Gate Array (FPGA) based hardware accelerator for 3D electromagnetic extraction, using Method of Moments (MoM). The method is applied to accelerate electrostatic parasitic capacitance extraction of multiple-nets in a Ball Grid Array (BGA) package. The acceleration is shown to be linearly scalable with FPGA resources and speed-ups over 10× against equivalent software implementation on a 2.4GHz Intel Core i5 processor is achieved using a Virtex-6 XC6VLX240T FPGA on Xilinx’s ML605 board with the implemented design operating at 200MHz clock frequency. This work was done in collaboration with Prof. Kuruvilla Varghese and his students from DESE. We presented our findings in a conference paper and have submitted a journal article. (b) Adaptive mesh refinement: 3D full-wave Method of Moments (MoM) based electromagnetic analysis is a popular means towards accurate solution of Maxwell's equations. However, the accurate solution of 3D full-wave MoM on an arbitrary mesh of a package-board structure does not guarantee accuracy, since the discretization may not be fine enough to capture spatial changes in the solution variable. At the same time, uniform over-meshing on the entire structure generates large number of solution variables and therefore requires an unnecessarily large matrix solution. We propose a suitable refinement criterion for MoM based electromagnetic package-board extraction. The advantages of this adaptive strategy are demonstrated from both accuracy and speed perspectives and our journal article has been accepted for publication in IEEE Trans. of CPMT.

Faculty Name: Dipanjan Gope


Spotlight: Multi-level modulation formats with coherent detection is studied and compared with the conventional IM/DD scheme. 4x4 MIMO and 6x6 MIMO system level perpeformance are studied with Few Mode Fiber as transmission medium.

Faculty Name: E.S.Shivaleela


Spotlight: A frequency discriminator circuit has been synthesized using RF all pass filters for real time Fourier transformation

Faculty Name: K.J.Vinoy


Spotlight: A spectral stochastic finite element method has been developed for handling uncertianties in electromagetic scattering problems

Faculty Name: K.J.Vinoy


Spotlight: Developed Novel Sparse Signal Recovery techniques.

Faculty Name: K.V.S.Hari


Spotlight: Transmit diversity for Spatial Modulation systems

Faculty Name: K.V.S.Hari


Spotlight: Developed fast eigen solution for solving homogeneous quadratic Minimization problems with constraints

Faculty Name: K.V.S.Hari


Spotlight: Developed channel estimation algorithms based on MBER framework for MIMO communication systems

Faculty Name: K.V.S.Hari


Spotlight: Developed Antenna Selection techniques for Spatial Modulation systems. Spatial Modulation schemes for low-complexity communication systemsDeveloped channel estimation algorithms based on MBER framework for MIMO communication systems. Developed fast eigen solution for solving homogeneous quadratic Minimization problems with constraints. Developed Novel Sparse Signal Recovery techniques using Fusion framework. Development of signal processing algorithms for indoor positioning systems using Inertial navigation systems.

Faculty Name: K.V.S.Hari


Spotlight: Proposed new performance criteria for evaluating the performance a new class of hybrid wireless sensor networks that consist of a mixture of energy harvesting sensor nodes and conventional unrechargeable battery-powered sensor nodes. Developed new, computationally efficient bounds for evaluating these criteria. Developed a new selection scheme that exploits power control to overcome collisions and combines it with a simple distributed, low feedback timer scheme to quickly and reliably find the best node among the set of available nodes. This has applications to a wireless systems such as vehicular networks, cooperative networks, sensor networks, and cellular systems that employ opportunistic node selection. Developed new and optimal transmit power adaptation policies for interference-constrained underlay cognitive radios for a variety of interference constraints imposed by spectrum regulators and for a variety of transmit power constraints motivated by practical power amplifier design issues. Proposed a new affine estimator for estimating the channel power gain from noisy estimates. The estimator has a simple, low complexity affine form and markedly outperforms conventional two-step channel amplitude-based estimation techniques at low signal-to-noise ratios.

Faculty Name: Neelesh B.Mehta


Spotlight: 1) India's largest telecom provider Bharti Airtel made an announcement in the last week of December 2014 that their customers will be charged more for voice-over-IP (VoIP) calls. Their justification was that they have made investments to the tune of Rs. 1.4 lakh crores, have paid an additional Rs. 0.5 lakh crore as government levies, and that offering VoIP services for free was not a tenable business model. They subsequently rolled back the plan for such charges citing reports that TRAI would issue a consultation paper shortly on the matter. We had already been studying such models of nonneutrality. We studied the effect a nonneutral network may have on end-users. We analysed two-sided market models and identified pure-strategy Nash equilibria, and further studied their relevance to existing broadband markets. 2) With applications to computational linguistics and semantic projection in mind, we studied the asymptotic optimality of belief propagation algorithms for finding the optimal edge cover in a random graph mean-field model. This provides an additional clue about the circumstances when belief propagation algorithms can be effective. 3) We studied the problem of distributed computation of a desired average on a network. We highlighted the issues with a classical asynchronous gossip scheme, proposed an alternative scheme based on reinforcement learning, and proved its convergence to the desired average. We proposed another reinforcement learning algorithm for distributed computation of the dominant eigenvector of a nonnegative matrix. This has applications to a wide variety of settings of current interest such as in ranking schemes, reputation networks, and principal component analysis. 4) We considered the problem of fair scheduling when mobiles can be strategic in reporting their true channel states. We investigated interesting equilibria that arise in such a network, and proposed mechanisms that were robust to such strategic behaviour.

Faculty Name: Rajesh Sundaresan


Spotlight: A fast fading interference channel with channel fade state known to all users was studied, treating interference at receivers as noise. The problem of state-dependent transmitter power allocation under average power constraint for each user was formulated as a stochastic game, assuming that the state distribution is known. Nash equilibria for the game were shown to be the solutions of an affine variational inequality. Sufficient conditions were found for a projection algorithm with regularisation to converge close to a Nash equilibrium, which are significantly weaker than conditions known previously for converging to Nash equilibria.

Faculty Name: Utpal Mukherji


Spotlight: Geographical forwarding in wireless sensor networks: In our earlier work, we had consider a light traffic setting where just one packet is traversing its way through the network, being locally optimally forwarded based on a local trade-off between link quality and progress. In our work during 2013, we extended this work to the situation where pairs of packets contend for relays. Markov stochastic game theory was used to derive locally optimal rules for the contending forwarder nodes to bid for a relay.

Faculty Name: Anurag Kumar


Spotlight: Competition over a user's timeline in online social networks: Online social networks are now being used more and more for disseminating content. In this situation, content creators can be viewed as contending for influence over the viewer. We have developed a stochastic model for a user's time-line which incorporates the following features: the level of influence that the content has, the decay of influence with time, the fact that the user scans the time-line from the top, thus paying less attention to items that low down on the screen or off the screen. We have also formulated the problem of competition for influence over the user.

Faculty Name: Anurag Kumar


Spotlight: Connection between discrete polymatroids, vertex graph colouring and Physical Layer Network Coding have been established. Also, it has been shown that space-time coding can be employed for X-interference network to lead to diversity gain apart from interference alignment. Fast-Decodable Space-Time Codes with large Coding Gain have been constructed for Multiple-input Double-output systems. Also, techniques for minimizing the complexity of fast sphere decoding of Space-Time Block Codes have been reported.

Faculty Name: B.Sundar Rajan


Spotlight: * Cognitive radio * Energy harvesting communications and computing * Compressive sensing and sparse Bayesian learning * Multiple antenna systems

Faculty Name: Chandra R.Murthy


Spotlight: Optical MIMO: We designed a mode converter, to convert a fundamental mode to higher order mode. Different modes act as distinct channels carrying distinct data which is used to enhance the capacity of fiber for eg two mode fiber. A two mode fiber model which supports a fundamental and a first-order mode is developed which can achieve higer capacity in comparison to a standard single mode fiber (SSMF). First-order mode however is four fold degenerate and can further be split into two more modes LP11a and LP11b thereby providing an additional degree of freedom for exploitation by mode division multiplexing (MDM) systems. Low loss mode multiplexer and de-multiplexer with low crosstalk are developed here to enable two mode multiplexing over the FMF. Wavelength division multiplexing is mounted on MDM system to further enhance the capacity of a wideband system.

Faculty Name: E.S.Shivaleela


Spotlight: A new project, on the design and characterization of CMOS based 20/60 GHz broadband transceiver components was started. The primary objective of this project, sponsored by DEITY, is application oriented research, which will involve the design of a critical RF component for Bharat Electronics Limited. In addition, the project will explore, using theoretical and experimental methods, various techniques of extending the range of operating frequencies for CMOS based RF circuits and systems. Ongoing work in CMOS based cognitive radio components, and the design of built-in-self-test (BIST) circuits for high frequency analog/RF applications has resulted in functional integrated circuits (chips) and a publication.

Faculty Name: Gaurab Banerjee


Spotlight: Significant research contributions during 2013: * In joint work with Prof. Manjunath Krishnapur (Math Dept., IISc), Eric Chan (CUHK) and Winston Fernandes (ME student), found thresholds for phase transitions in the Bethe approximation to the PML estimate. * In an ongoing collaboration with Manuj Mukherjee (PhD student) and Yogesh Sankarasubramaniam, we are investigating the type of interactive public communication necessary to generate an optimal-rate secret key from a source of correlated randomness. Some of our results have been submitted to ISIT 2014. * In an ongoing collaboration with Prof. Sidharth Jaggi's research group at the Chinese University of Hong Kong (CUHK), we are trying to obtain performance guarantees for the generalized belief propagation algorithm when used in the problem of counting constrained binary arrays. * In joint work with Shashank V. (PhD student), we are trying to reduce the complexity of lattice coding-based algorithms for reliable and secure computation in the setting of a bidirectional relay with Gaussian noise. In particular, we are investigating the use of nested lattice codes using Low-Density Construction-A (LDA) lattices.

Faculty Name: Navin Kashyap


Spotlight: Our focus over the past year has been on coding for distributed storage networks. The largest family of constructions for optimal regenerating codes have originated from our group here at IISc. We now have three journal publications in this area in the IEEE Transactions on Information Theory which cumulatively have over 200 GS citations. Under a faculty fellowship grant from NetApp, we are jointly pursuing implementation of some of our code constructions. We have also contributed to the development of the theory of codes with locality by extending their definition and then designing optimal codes that combine the properties of both regenrating codes as well as codes with locality. Our work in wireless sensor networks has been concerned with the simulation and hardware platform design of a PIR-sensor intended for intrusion detection. Here we have been funded under a joint NSF-DeITY initiative called PC3. Our work on distributed function computation has just been accepted to as special issue of JSAC dealing with In-Network Computation.

Faculty Name: P.Vijay Kumar


Spotlight: We studied the problem of distributed computation of a desired average on a network. We highlighted the issues with a classical asynchronous gossip scheme, proposed an alternative scheme based on reinforcement learning, and proved its convergence to the desired average. We proposed another reinforcement learning algorithm for distributed computation of the dominant eigenvector of a nonnegative matrix. This has applications to a wide variety of settings of current interest such as in ranking schemes, reputation networks, and principal component analysis.

Faculty Name: Rajesh Sundaresan


Spotlight: The optimal tradeoff between average transmit power and average data queueing delay at the transmitter, for a fading point-to-point link with perfect channel state information at the transmitter, was comprehensively analysed through geometric bounds on the stationary probability distribution of the queue length.

Faculty Name: Utpal Mukherji


Spotlight: High Speed TCP Protocols: TCP cubic and TCP compound were studied and their analysis in multirouter, multi-TCP environment was provided. Algorithms for Spectrum Sensing in a wide-band environment when observations are received in compressed domain were developed. Also, universal source coding techniques were used for sensing in nonparametric sequential setup. Cooperative ARQ protocols were developed to enhance throughput and fairness and ensure Quality of Service in a WLAN network. Efficient power allocation algorithms for interference channels were developed.

Faculty Name: Vinod Sharma


Spotlight: Ubiquitous Learning- Adaptive Web based Ubiquitous Learning tends to be more promising as it offers unlimited level of personalization and satisfaction of goals and preferences of particular student. Moreover, based on knowledge level, content delivery adapts itself to suit the student's requirement during learning process. Domain Model uses a tree with multiple links. Each link weight has been derived from the conceptual relations, among the concepts of the subject. The designed modular tree with multiple conceptual links was very useful in guiding diverse students through the course-ware.

Faculty Name: Anandi Giridharan


Spotlight: SmartConnect: Design algorithms and an on-line design and deployment tool for wireless sensor networks.

Faculty Name: Anurag Kumar


Spotlight: Geographic forwarding in sleep-wake cycling wireless relay networks that interconnect sensor nodes to a sink. Our approach is to formulate local forwarding as optimal sequential decision problems. The reward structure accounts for progress towards the sink, the delay in selecting a forwarding relay, and the power required to forward to the relay. The formulations give rise to variants of asset selling problems.

Faculty Name: Anurag Kumar


Spotlight: Optimal impromptu (or place-and-go) deployment of a wireless relay networks for sensor interconnection. The problem is modeled as a sequential decision problem, leading to a constrained MDP that is solved to obtain structural results.

Faculty Name: Anurag Kumar


Spotlight: Optimal epidemic forwarding of content in a mobile opportunistic network where there are nodes that are interested in the content and those that only serve as relays. We model the spread of interest via a social network model. The problem is to tradeoff the cost of forwarding to a relay against the benefit of more mobiles to forward the content. A fluid limit is taken, leading to a controlled ODE, whose solution yields a distributed control.

Faculty Name: Anurag Kumar


Spotlight: Physical Layer Network Coding methods for the K-user Multiple Access Relay Channel have been found and are to be presented in IEEE WCNC conference in Shanghai. Also, it has been shown that space-time coding can be employed for two-way wireless information exchange which perform better compared to the traditional techniques in Rician fading channels. In the area of network coding the relationship between matroids and network coding have been extended by way of giving constructions for Matroidal Error Correcting Networks, In the area of space-time coding it is shown that for any space-time code the complexity of maximum-likelihood decoding with GDL (generalized distributive law) is always less than or equal to the complexity with CML (conditional ML decoding).

Faculty Name: B.Sundar Rajan


Spotlight: * Cognitive radio * Energy harvesting communications and computing * Compressive sensing and sparse Bayesian learning * Scheduling and resource allocation/optimization

Faculty Name: Chandra R.Murthy


Spotlight: In the last 1 year we have conducted research and development work in different aspects of Computational Electromagnetics: Shared-Memory Parallelization of fast solver algorithms: An efficient parallelization algorithm for the Fast Multipole Method was presented which aims to alleviate the parallelization bottleneck arising from lower job-count closer to root levels. An electrostatic problem of 12 million non-uniformly distributed mesh elements is solved with 80-85% parallel efficiency in matrix setup and matrix-vector product using 60GB and 16 threads on shared memory architecture. Distributed-Memory Parallelization of Method of Moments: Volume integral equation (VIE) for electromagnetics was solved for homogeneous and inhomogeneous dielectric objects for radar cross section (RCS) calculation in a highly parallel environment using parallel numerical library ScaLAPACK on IBMÂ’s distributed-memory supercomputer BlueGene/L by different number of processors to compare the speedup. RF Imaging: We have initiated research in fast direct electromagnetic solvers for adjoint-variable-method (AVM) based sensitivity matrix computations towards imaging applications. Different reconstruction algorithms like Tikonov regularization, Landweber iterations, and Linear Back Projection based iterative methods have been applied to reconstruct the image in Electrical Capacitance Tomography (ECT) which is applicable in detecting impurities in fluid flow exploiting their dielectric variations. In the next stage the same algorithms will be tailored for use in RF imaging for non-invasive, early detection and monitoring of breast cancer.

Faculty Name: Dipanjan Gope


Spotlight: Designed and demonstrated a CWDM digital distribution network of 1x64 split Passive optical network. We have also designed and developed a opticallly amplified PON with 4 wavelengths.

Faculty Name: E.S.Shivaleela


Spotlight: Development of signal processing algorithms for indoor positioning systems using Inertial navigation systems. Detection and classification of vehicles using magnetometer data. Fusion Framework for Greedy Pursuit Algorithms. Spatial Modulation schemes for low-complexity communication systems

Faculty Name: K.V.S.Hari


Spotlight: Development of signal processing algorithms for indoor positioning systems using Inertial navigation systems. Detection and classification of vehicles using magnetometer data. Fusion Framework for Greedy Pursuit Algorithms. Spatial Modulation schemes for low-complexity communication systems

Faculty Name: K.V.S.Hari


Spotlight: ML algorithms for joint estimation of channel and impairments.

Faculty Name: K.V.S.Hari


Spotlight: Significant research contributions during 2012: * In joint work with Prof. Andrew Thangaraj (IIT Madras) determined an exact expression for the treewidth of MDS and Reed-Muller codes * In joint work with Himanshu Tyagi (University of Maryland), Yogesh Sankarasubramaniam (HP Labs, Bangalore) and Manuj Mukherjee (PhD student), formulated and analyzed a new model of fault-tolerant secret key generation. * In joint work with Prof. Andrew Thangaraj (IITM) and Shashank V. (PhD student), determined achievable rates for reliable and perfectly secure computation in the setting of a bidirectional relay with Gaussian noise. * In joint work with K.M. Shivkumar (PhD student), formulated and analyzed a 2-dimensional error model for data storage on a granular magnetic medium. * In joint work with Winston Fernandes (ME student), found a necessary and sufficient condition for a uniform distribution to be a local maximum for the optimization problem of pattern maximum likelihood * In joint work with Prof. Gilles Zemor, University of Bordeaux, determined new upper bounds on the cardinality of grain-correcting codes.

Faculty Name: Navin Kashyap


Spotlight: Our research group interests are exploring and designing optimum networking solutions and protocols for MANETs, ubiquitous learning, ubiquitous, pervasive and social networks. These works have been undertaken in mobile & ubiquitous computing. In MANETs, we have solved the following prominent issues like node monitoring, resource management and routing. By using agent technology which uses both static and mobile agents, when compared with other methods, this method reduces the network response time, bandwidth usage and the energy consumption with varying network load conditions. The ubiquitous network, which provides uninterrupted connectivity to the users anytime, anywhere, any device on any network technology, is the fundamental requirement for designing the ubiquitous applications. We have undertaken the task of designing solutions to some of the existing issues in ubiquitous network like the ubiquitous application node monitoring, optimized resource allocation, routing and reliable data delivery for ubiquitious training by using agents. Pervasive computing is a shift in computing paradigm, where we have designed a policy based user privacy scheme by collecting the context information of the users and their environment. Designing a generic actor model for social networks has been carried out and tested over the agriculture social network, and well appreciated by the research community. Industriay initiated projects successfully include ubiquitous traffic management system, PETUbiNet simulator and ubiquitous museum guide system.

Faculty Name: P.Venkataram


Spotlight: We modeled a search problem in visual neuroscience as an optimal control problem with stopping. This enabled us to provide a decision theoretic explanation for the strong correlation between delay in distinguishing two images and the similarity of the two images. With applications to computational linguistics and semantic projection in mind, we studied the asymptotic optimality of belief propagation algorithms for finding the optimal edge cover in a random graph mean-field model. This provides an additional clue about the circumstances when belief propagation algorithms can be effective. We also proposed a decentralised and asymptotically optimal push-pull procedure for multicasting in a wireless network. This would be of great use for simultaneous and live viewing of videos of important events on the web.

Faculty Name: Rajesh Sundaresan


Spotlight: Minimum average delay was characterized for a single-server iid-batch-arrival batch-service queue, with batch service time, a deterministic function of batch size. Heuristic batch size selection policies were proposed for which the asymptotic growth rate of average delay matches that of minimum average delay. The analysis was applied to the study of code rate selection policies for a point to point link.

Faculty Name: Utpal Mukherji


Spotlight: Performance analysis of TCP over IEEE 802.11 CSMA/CA while incorporating the effect of MAC level packet dropping due to excessive channel induced packet errors.

Faculty Name: Anurag Kumar


Spotlight: Performance analysis of multihop CSMA/CA networks arising in Zigbee sensor networking.

Faculty Name: Anurag Kumar


Spotlight: Adaptive Wireless Network Coding techniques have been found for two-way and three-way wireless information exchange through a relay. This problem has been shown to be closely related to existence of combinatorial structures called Latin Squares. In the area of space-time codes, low ML decoding complexity codes have been found using classical codes over Klein groups. Moreover, Collocated and Distributed Space-Time Codes s with Partial Interference Cancellation Decoding have been reported. In a direction of research, a low ML-decoding Complexity, Full-diversity, Full-rate MIMO Precoder have been found. All these works have appeared in IEEE Transactions on Wireless Communication/ Information Theory /Signal Processing.

Faculty Name: B.Sundar Rajan


Spotlight: Developed new switching techniques for low power SAR ADCs. Demonstrated that via a test chip in 130nm CMOS Developed new diode-equivalent circuit which is better than state of the art and can be used for energy scavenging circuits in ultra low power applications. Developed and demonstrated a new, high accuracy, arbitrary delay synthesis technique which will be useful in applications like integrated circuit test, ranging etc.

Faculty Name: Bharadwaj Amrutur


Spotlight: * Cognitive radio * Energy harvesting communications and computing * Compressive sensing and sparse Bayesian learning * Scheduling and resource allocation/optimization

Faculty Name: Chandra R.Murthy


Spotlight: Design and development of a prototype large split wavelength division multiplexing optical network for distribution of RF and digital signals for RADAR applications

Faculty Name: E.S.Shivaleela


Spotlight: Efficient routing in translucent networks to achieve efficiency close to transparent networks using minimal network resources.

Faculty Name: E.S.Shivaleela


Spotlight: Performance evaluation of Optical CDMA networks with !-D and 2-D codes with correlation and chip level receivers, for access networks

Faculty Name: E.S.Shivaleela


Spotlight: Developed channel estimation algorithms using the Minimum BER criterion and BPSK signals which shows improvement in BER compared to MMSE estimation and detection. Lox complexity adaptive algorithms for channel shortening filters in OFDM systems Developed novel estimation technique of channel shortening filters for MIMO OFDM systems Developed modification to CSTSK scheme for MIMO communciation with Field-Extension Dispersion Matrices improving the coding gain and the BER Developed new waveforms for MIMO Radar using Costas array based technique Developed capacity bounds for MIMO-X Relays in Interference Alignment framework Developed hierarchical Interference Alignment techniques in a MIMO Interference network.

Faculty Name: K.V.S.Hari


Spotlight: Developed Field-extension Dispersion matrices for coherent space-time-shift keying (CSTSK) systems to improve coding gain and improved BER performance

Faculty Name: K.V.S.Hari


Spotlight: Developed low complexity adaptive algorithms for channel shortening in OFDM systems

Faculty Name: K.V.S.Hari


Spotlight: 1. Developed a comprehensive performance analysis of feedback reduction techniques employed by the Long Term Evolution (LTE) standard. 2. Developed a robust version of the splitting-based selection algorithm to handle uncertainty in number of nodes in the system and its robustness to uncertainty in system parameters. 3. Highlighted an inefficiency that arises in training for antenna selection and proposed new techniques to optimally tackle this inefficiency 4. Showed that buffering at relays can significantly improve the performance of a rateless code based multi-relay cooperative communications system.

Faculty Name: Neelesh B.Mehta


Spotlight: We proposed and analysed delay and energy optimal epidemic relaying protocols for delivery of packets in delay-tolerant networks. We continued our study on the effect a nonneutral network may have on end-users. We analysed two-sided market models and identified pure-strategy Nash equilibria, and further studied their relevance to existing broadband markets. We came up with a generalized family of relative entropies and explored their geometric properties. These were done with the aim of providing an axiomatic framework for the theory of inference. We studied the behaviour of wireless local area networks, analysed complexity reduction heuristics, and provided a justification for their accuracy.

Faculty Name: Rajesh Sundaresan


Spotlight: Design of mux/Demux using photonic band gap materials around 1550 nm used for optical communications. Fabrication of this optical device is in progress.

Faculty Name: T.Badrinarayana


Spotlight: 1. Compressive sensing for speech and music 2. Multi-sensor speech processing for CASA 3. Non-linguistic approach to language identification 4. Stochastic resonance in auditory signal processing 5. Time-frequency models of timbre in audio processing 6. Human-robot communication 7. Audio classification and retrieval

Faculty Name: T.V.Sreenivas


Spotlight: 1. Change detection in wireless sensor networks: (a) obtained the optimal stopping policy for change detection in a collocated sensor network in which the measurements arrive over a random access network and the decision algorithm processes the arriving measurements in a network aware manner; (b) Obtained optimal change detection rules for the detection of transient changes. 2. Geographical forwarding: Extended the optimal forwarding work we did earlier to the case in which the forwarding node does not know the number of forwarding neighbours. This is studied as a partially observable MDP. 3. Information dissemination in social networks: conducted an analytical study of Kempe, Kleinberg, Tardos's linear threshold model for influence in a social network. Provided recursive analytical expressions, and a fluid limit for the trajectory of the activation process.

Faculty Name: Anurag Kumar


Spotlight: Constellation constrained capacity (CCC) for basic multiuser communication (with finite complex constellation inputs) have been studied for basic channels like, Multiple Access Channel (MAC), Broadcast Channel and Interference Channel. For the 2-user Gaussian MAC, for certain cases it is shown that TCM (Trellis Coded Modulation) can approach the CCC. In the context of multiple access, (orhogonal vs non-orthogonal) results that are not seen when Gaussian inputs are considered are reported. These results are to appear in the March 2011 issue of IEEE Transactions on Information Theory.

Faculty Name: B.Sundar Rajan


Spotlight: Demonstrated a low cost, battery operated pollution monitoring system based on GSM connectivity. Now we are working on a large scale deployment of the same in the city. Developed and demonstrated some new techniques for measuring very small delays. We have applied for a couple of patents for the same. This technique has many applications in the chip design area. Led the development of a multi-core simulator which allows for flexible experimentation. It is now being used by many graduate students for their research work. Developed techniques for modeling of delay and leakage power of logic gates, which is very amenable to statistcial analysis. This will be very useful for Chip design and development by the semiconductor industry.

Faculty Name: Bharadwaj Amrutur


Spotlight: * Source coding and joint source-channel coding * MIMO/OFDM channel estimation * Cognitive radio * Machine learning

Faculty Name: Chandra R.Murthy


Spotlight: Evaluating different protocols for optical CDMA networks using 1-D codes and also 2-D wavelength/time codesfor access networks. Developing Fiber distribution network, where large splitting of signals is required such as passive optical networks, transmission of control signals to antenna array etc,, using wavelength division multiplexing technique. The advantage here, is the replacement of conventional (many) co-axial cables by (one) optical fiber, reduction of foot-print weight and also EMI.

Faculty Name: E.S.Shivaleela


Spotlight: Started a project on integrated cognitive radio circuits for the 1-10 GHz frequency range. Cognitive radios for strategic applications are being explored with emphasis on broadband RF front ends, fast frequency synthesis, efficient RF spectrum sensing and reconfigurable A/D converters. Proposals for external funding in the pipeline. 4 PhD students working on the project.

Faculty Name: Gaurab Banerjee


Spotlight: A new ultrawideband antenna developed. Currently investigating its applications in areas such as ultrawideband microwave imaging.

Faculty Name: K.J.Vinoy


Spotlight: RF parts of an integrated phased array is implemented on a microwave laminate using the fabrication approach developed in-house. The main component of this array is the meso-scale electrostatic phased shifter on laminate (MEPL).

Faculty Name: K.J.Vinoy


Spotlight: Developed subspace methods using the Khatri-Rao product of matrices to estimate multiple CFOs in a Frequency reuse-1 MIMO-OFDM/OFDMA Wireless system. Studied the performance on acoustic source localization techniques using microphone arrays. Studied algorithms for detecting vehicles using magnetometer sensors. Studied the diversity and complexity of MIMO Zero-padded oFDM systems. Obtained novel properties of autocorrelation sequences base don the LSF model. Studied the tracking algorithms in a wireless sensor network using Kalman Filtering principles. Studied the joint MMSE receivers using FIR precoding and equalization. techniques

Faculty Name: K.V.S.Hari


Spotlight: 1. Showed how energy harvesting functionality in relays is a promising tool to enhance cooperative wireless networks. 2. Developed novel energy-efficient training protocol and a new optimal antenna selection rules for antenna selection in time-varying wireless environments. 3. Developed an optimal timer-based, low feedback distributed selection algorithm to select the best node among a given set of nodes. The algorithm has wide applications in cooperative networks, cellular systems, vehicular ad hoc networks, and wireless sensor networks.

Faculty Name: Neelesh B.Mehta


Spotlight: Design of context aware system, issues related to ubiquitous multimedia, social network, personalization of services in u-computing, seamless connectivity in U-computing, Device adaptability.

Faculty Name: P.Venkataram


Spotlight: We provided a finite-state link reversal algorithm to render a destination-disoriented directed acyclic graph into a destination-oriented one. This has applications in keeping a wireless sensor network in a prepared state even when a few nodes can fail. We provided fair scheduling algorithms in cellular systems that were robust to strategic behaviour by noncooperating mobiles. We proposed and analysed delay and energy optimal two-hop relaying protocols for delivery of packets in delay-tolerant networks. We initiated a study on the effect a nonneutral network may have on end-users. We proposed two-sided market models, analysed the ensuing equilibria, and studied their relevance to existing broadband markets.

Faculty Name: Rajesh Sundaresan


Spotlight: Design of ring resonator based optical filter at 1550 nanometer wavelength for WDM networks, and the fabrication of the same device is in progress. Since the dimensions of these devices are small, effort are put in to procure the characterization equipment. In future work, I propose to investigate application of photonic bandgap devices to 4 channel multiplexer an demultiplexer used for network applications.

Faculty Name: T.Badrinarayana


Spotlight: Characterisation of average delay - average error probability tradeoff for point-to-point channels using bound on error probability as function of message batch size and batch codeword length..

Faculty Name: Utpal Mukherji


Spotlight: Cognitive Radio: Dual CUSUM algorithm developed earlier by us was extended to the scenario when the Primary user OFDM. This involved addressing new issues in handling time and frequency offset and IQ imbalance in OFDM systems. Next we considered the problem when the primary user OFDMA. In this case the previous algorithms are not suitable to identify the different channels which are free at a given time. For this we have developed new algorithms using compressive sensing. Another problem addressed is the modelling and performance analysis of real time and data traffic carried by the Cognitive radio users which use the channel when the primary is not using it. Energy Harvesting sensor networks: Earlier we had modelled an energy harvesting sensor node and obtained delay optimal and throughput optimal energy management policies. In this year we provide information theoretic fundamental limits on communication rate for such nodes.

Faculty Name: Vinod Sharma


Spotlight: Our recent research on multi-antenna wireless communications has established that large MIMO lsystems with tens of transmit/receive antennas can be practical. High spectral efficiencies (tens to hundreds of bps/Hz) can be acieved can be achieved using such large MIMO systems

Faculty Name: A.Chockalingam


Spotlight: 1. Geographical forwarding in wireless sensor networks: developed a tunable delay-progress optimal forwarding scheme in the situation in which nodes are sleep-wake cycling 2. Performance of wireless local area networks: (a) provided an analytical performance model for the power save mode in WLANs, (b) developed an analytical model for infrastructure WLANs carring short-lived file transfers (as would arise in web browsing). 3. Change detection in sensor networks: obtained the optimal stopping policy for change detection in a collocated sensor network in which the measurements are sent to the fusion centre over a random access network, and fusion centre processes measurements as they arrive but in time-causal order. 4. Developed a game-theoretic model for power control in multichannel CDMA cellular systems.

Faculty Name: Anurag Kumar


Spotlight: (1) Space-Time codes for 2x2 and 4x2 MIMO systems with lower ML decoding complexity than the Golden code and the DjABBA code have been constructed and published in the Journal of Slected Topics in Signal Processing, Dec' 2009 issue.

Faculty Name: B.Sundar Rajan


Spotlight: At the VLSI Circuits and Systems lab in the ECE Department, we >have been working on developing techniques for designing low power mixed >signal integrated circuits and systems for applications like wireless >sensor networks, >neuro-electronic interfaces, low power video encoders etc. As an >example, we have developed a >

Faculty Name: Bharadwaj Amrutur


Spotlight: * Source coding and joint source-channel coding * MIMO/OFDM channel estimation * Cognitive radio * Machine learning

Faculty Name: Chandra R.Murthy


Spotlight: Digital processing of biomedical signals like EEG, HRV and pulse has been carried out. Functional connectivity analysis of brain regions has been made using graph theoretic approach. A new method of estimating the fractal dimension of signals has been proposed and then applied successfully to schizophrenic EEG data. Chaos characterization and recurrent quantification analysis has been made on HRV data. Pulse signals have been recorded on the wrist and analyzed using various DSP techniques.

Faculty Name: D. Narayana Dutt


Spotlight: Optical CDMA plays an important role in the self-routing of packets, in all-optical netwoks. We sed this technique in a fixed routing network which greatly simplifies the node complexity and also the delay in forwarding, as there is no optical-electrical-optical conversion.

Faculty Name: E.S.Shivaleela


Spotlight: In the area of microwave antennas, a multi-functional printed antenna was developed under a sponsored research activity. One of these antennas consists of two U-shaped slots printed on a microstrip patch on a customized substrate, operates at two wide-bands centered around 2.1 GHz and 5.5GHz and cater to wireless applications including GSM, WiMAX, UMTS, WLAN and HyperLAN with good overall performance. The measured signal strength from this antenna is symmetrical in one planes, at both operational ranges.

Faculty Name: K.J.Vinoy


Spotlight: A low cost electrostatically actuated phase shifter is demonstrated for microwave applications. This is fabricated with materials and fabrication processes commonly used in microwave planar circuits. Work is currently underway to build a complete phased array antenna with four elements and these phase shifters.

Faculty Name: K.J.Vinoy


Spotlight: Developed a new representation of a complex finite length sequence using the Line Spectral Frequency (LSF) model. This novel LSF representation provides new insights into the relation between the phase and magnitude spectra of sequences leading to very useful DSP applications One such application is the development of novel DSP filter structures which are based only on Notch filters. These structures are less sensitive to co-efficient quantization effects than conventional structures. A new representation of complex numbers called the Line Spectral representation is proposed. This representation is different from the polar and cartesian representations and is used to build novel arithmetic modules for DSP applications.

Faculty Name: K.V.S.Hari


Spotlight: 1. Developed a fast dual power multiple access algorithm that has a stable throughput of 0.793 packets/slot, which to the best of our knowledge is the highest achievable by multiple access protocols today. 2. Developed energy-efficient routing techniques for cooperative wireless networks. 3. Developed relay load balancing techniques for cooperative wireless networks that use rateless codes that reduce the delays faced by packets in queued cooperative networks by more than 50%, while incurring a negligible reduction in throughput. 4. Initiated work on communication link performance characterization and optimization for energy harvesting sensor networks.

Faculty Name: Neelesh B.Mehta


Spotlight: An Authentication Scheme for Ubiquitous Public Transport Security and Emergency Management System

Faculty Name: P.Venkataram


Spotlight: Context aware based web access system for Museum

Faculty Name: P.Venkataram


Spotlight: A Multiagent Based context aware system for Ubiquitous Public Transport security and emergency Management system

Faculty Name: P.Venkataram


Spotlight: Transactions based dynamic security and authentication scheme for mobile communications environment, using cognitive agents.

Faculty Name: P.Venkataram


Spotlight: We obtained explict code constructions which minimize the amount of data downloaded to repair a failed node in a distributed storage setting. Prior to this work, only existence of certain codes had been proved, and the approaches considered in the literature required very high field sizes and were practically not feasible. Our codes require very low field sizes, are of low complexity and require low maintainence, all making them very practical. Various codes were presented at Allerton conference in 2009, ITW Cairo in 2010 and some will also be presented at the National Conference on Communications in Jan. 2010. Our work has been well recieved by the community. Though it is a very recent work, it has been cited in works such as [1] and have also been used by Y. Wu [2] to construct codes for a slightly different setting for distributed storage. Also, Changho et al.[3] have proved certain additional properties of our codes, thereby closing a major open problem.

Faculty Name: P.Vijay Kumar


Spotlight: Provided a convex optimisation framework to certain sequence design problems in wireless uplink with applications to a large class of optimisation problems; Obtained cross-layer policies for queue management and control under conditions of uncertain information; Identified decentralised flow optimisation strategies in a network; Designed robust mechanisms to induced cooperative behaviour in wireless networks; Designed budget-balanced mechanisms for allocation of divisible goods among agents; Designed robust link reversal algorithms for a wireless sensor network.

Faculty Name: Rajesh Sundaresan


Spotlight: Design of ring resonator based optical filter at 1550 nanometer for WDM networks,fabrication and characterization of the same device is under progress. Work is under progress on the design and fabrication of Photonic bandgap based optical waveguides and resonators used for network application.

Faculty Name: T.Badrinarayana


Spotlight: Average-delay optimal policies for the point-to-point channel- Average-delay optimal scheduling of messages arriving to the transmitter of a point-to-point channel was studied. A batch-service queueing model for the communication scheme was considered, with service time an affine function of the batch size. The question of delay optimality was addressed within a semi-Markov decision-theoretic framework. Approximations to the average-delay optimal policy were obtained.

Faculty Name: Utpal Mukherji


Spotlight: Distributed Event Detection in Wireless Sensor Networks ; Studied three distributed event detection schemes based on CUSUM, in a large network setting, and establishment order optimality of two them in a multihypothesis minimax optimal change detection setting. Game Theoretic Approaches for Resource Allocation in Wireless Network ; 1. Formulated the problem of cellular operator cooperation as a coalition game and established the nonemptiness of the core ; 2. Studied optimal SNIR based association of mobiles with base stations in 1-dimensional space in a noncooperative game setting 3. Studied uplink power control in a noncooperative game setting, in a situation in which base stations use different channels, and provided a pricing based approach for achieving a Pareto optimal power control. IEEE 802.11 Wireless Local Area Network ; 1. Provided an analytical model for unfairness between TCP controlled uplink and downlink file transfers in an infrastructure WLAN. 2. Provided an analytical model for TCP throughouts in a system of interfering infrastructure WLANs, under a pairwise binary dependence model. 3. Completed the first phase of development of Wireless LAN Manager (WM) a system for centralized management of QoS in infrastructure WLANs.

Faculty Name: Anurag Kumar


Spotlight: (1) A low-complexity detector, called Likelihood Ascent Search (LAS) detector, has been obtained that can be used to decode large dimensional space-time block codes, in collaboration with Prof. Chockalingam.

Faculty Name: B.Sundar Rajan


Spotlight: Working on routing in optical burst switching networks with physical layer impairment awareness

Faculty Name: E.S.Shivaleela


Spotlight: A multi-layer planar approach to implement high surface impedance and suppression of surface wave propagation proposed

Faculty Name: K.J.Vinoy


Spotlight: A wideband antenna with circular polarization designed

Faculty Name: K.J.Vinoy


Spotlight: A planar antenna designed for various wireless bands below 6GHz designed

Faculty Name: K.J.Vinoy


Spotlight: An approach to miniaturize microstrip filters using the photonic bandgap concept proposed

Faculty Name: K.J.Vinoy


Spotlight: Line Spectral Frequency representation of complex sequences and novel FIR filter structures based on LSFs.

Faculty Name: K.V.S.Hari


Spotlight: 1. Developed several fast, distributed, and scalable algorithms for selecting the best instantaneous relays in cooperative communication systems; 2. Propose novel weighted selection criteria to improve antenna selection performance in next generation wireless communication systems; 3. Developed an accurate lognormal model for modeling interference in uplinks of CDMA systems; 4. Investigated precoding schemes to tackle a fundamental asymmetry in channel knowledge in wireless relay systems in infrastructure-based systems

Faculty Name: Neelesh B.Mehta


Spotlight: Wireless sensor networks: identified algorithms for computation of histogram of observations, designed algorithms for synchronisation. Identified strong connections between guessing, secrecy, and compression via applications of large deviations theory. Identified equilibria in user-association and base station placement games. Studied distributed power control algorithms in wireless systems from a game theoretic perspective. Designed mechanisms to induce cooperative behaviour in wireless networks.

Faculty Name: Rajesh Sundaresan


Spotlight: Micro ring resolnators, various configurations are modeled and simulated. These find applications in WDM mux and demux. Inhomogenous and nonlinear optical effects in DWDM optical communication systems are explored. New techniques for QOS aware quoramcasting for OBS networks are explored.

Faculty Name: T.Srinivas


Spotlight: Scheduling for Proportional Throughput Differentiation: We analysed a scheduling policy which balances weighted cumulative numbers of packet departures at a multiplexer queue, for the basic case of two traffic classes, and showed that the average throughputs achieved by the classes are proportional to their desired throughputs; this represents an improvement compared with Weighted Round Robin scheduling when class window sizes are small or class round trip times are large. Optimal Energy Management Policies for an Energy Harvesting Sensor Node: We obtained energy management policies for an energy harvesting sensor node that support maximum possible throughput or minimise mean packet delay while operating for indefinitely long periods within the harvested power budget of the node.

Faculty Name: Utpal Mukherji


Spotlight: Cooperative Communication: Showed that for cooperative communication relay networks with rateless codes the throughput can be substantially increased if the relays have buffers to store packets. Also obtained load balancing techniques to reduce queueing delays in such systems. Developed distributed cooperative algorithms for spectrum sensing which are highly energy efficient and perform better than other known algorithms. Also these algorithms do not suffer from SNR wall problem in the presence of uncertainties in noise and primary transmit power. Also developed efficient energy management policies, MAC protocols and joint power, routing and scheduling policies for energy harvesting sensor networks.

Faculty Name: Vinod Sharma


Spotlight: Wireless Sensor Networks: design and analysis of algorithms for self-organisation, localisation, distributed detection, and function computation. On the problem of sequential change detection we introduced the aspect of network delay for the case in which observations flow over a network; further, we introduced a cost for making measurements, thus providing a technique for turning sensors on and off based on observations. We developed a distributed heuristic algorithm for the construction of the critical geometric graph on randomly located nodes, with applications to self-organisation.

Faculty Name: Anurag Kumar


Spotlight: (2) Several interesting results in the topic of low decoding complexity distributed space-time block codes (DSTBCs) have been obtained. Specifically, Algebraic DSTBCs with low decoding complexity for non-coherent communication and coordinate interleaved designs for two-antenna-relays networks have been accepted for publication in IEEE Trans. on Wireless Communication. (3) New set of square complex orthogonal designs with low PAPR (Peak to average power ratio) and signalling complexity have been found (To appear in IEEE Trans. Wireless Communication.) (4) A class of algebraic space time block codes for colocated MIMO communication that are optimal for MMSE detection have been found. (To appear in IEEE Trans. on Wireless Comm.)

Faculty Name: B.Sundar Rajan


Spotlight: Succesful collaboration with Texas Instruments to design and tapeout a chip in their advanced 65nm CMOS Process

Faculty Name: Bharadwaj Amrutur


Spotlight: * Source coding and joint source-channel coding * MIMO/OFDM channel estimation * Cognitive radio * Machine learning

Faculty Name: Chandra R.Murthy


Spotlight: • Chief software architect and solver implementation for nWave, nApex at Nimbic Inc a world-wide provider of 3D full-wave electromagnetic solution

Faculty Name: Dipanjan Gope


Spotlight: Proposed and analyzed a new family of wavelength/time codes for Fiber-Optic CDMA networks. As these codes have minimal cross correlation values, they perform better in a bursty traffic environment

Faculty Name: E.S.Shivaleela


Spotlight: Designed dual frequency antennas using fractal shaped geometries for wireless bands

Faculty Name: K.J.Vinoy


Spotlight: New equivalent circuit models developed for miniaturized bandpass filters

Faculty Name: K.J.Vinoy


Spotlight: Novel approaches for FIR filter structures, Novel Communication System Design

Faculty Name: K.V.S.Hari


Spotlight: Transaction based security and authentication system for mobile communication. Web bbased Education. Seamless connectivity in ubiquitous system. Network management. Context aware ubiquitous commerce. Ubiquitous simulators and ubiquitous operating system. Datastorage and access in ubiquitous computing environment. Ubiquitous Computing TestBed.

Faculty Name: P.Venkataram


Spotlight: Space-time Codes: The focus of our research progressed here from optimal space-time code construction for point-to-point communication to distributed space-time codes for cooperative relay communication in wireless networks: 1) Paper constructing perfect space-time codes for point-to-point communication for all numbers of antennas at transmitter and receiver appeared (Nov. 2007 Trans-IT) 2) construction of low-complexity space-time codes for cooperative communication appeared in the special issue of JSAC on cooperative communication and networking (Feb 2007) 3) codes for optical CDMA based on cyclic difference sets appeared ( May 2007 issue of IEE Trans-IT) 4) manuscript on optimal distributed space-time codes for two-hop networks has been accepted for publication pending minor revision 5) manuscript on optimal MIMO-ARQ space-time codes was accepted pending revision; revision has been submitted Conference Organized: 6) in a preprint we provide optimal constructions for the dynamic decode and forward protocol as well as DMT optimal codes for the MIMO OFDM channel 7) most recently we have been successful in characterizing the DMT of multi-hop networks and will soon submit to Trans-IT Low-Correlation Sequences We provided the first constructions for codes for CDMA for the popular QAM signal constellation; the sequences constructed permit variable rte signalling on the reverse link of a CDMA system and in this connection an Indian patent has been filed The journal paper is scheduled to appear in the Feb 2008 issue of the Trans IT. Sensor Networks: Intrusion detection has been considered for a hexagonal sensor grid; a signal-space viewpoint was introduced that is useful in performance analysis; the problem of minimizing the number of bits transmitted is under study and some preliminary results have been obtained; belief propogation has been used to yield a useful localization algorithm Conference Organization: a) General Chair, 17th Symposium on Applied Algebra, Algebraic Algorithms and Error-Correcting Codes (AAECC-17) , Dec. 17-20, 2007, Bangalore. b) Technical Program Co-Chair: IEEE Information Theory Workshop, July 1-6, 2007 Bergen Norway Invited Talks a) UCSD-ITA Workshop b)NCC 2007, IIT Kanpur c) Sequences, subsequences and their consequences, May 30-31, 2007, Los Angeles d) Tor Helleseth 60th Birthday symposium, Bergen, Norway, Nov. 21-22, 2007 e) invited to participate in the Oberwolfach Workshop on Coding Theory (December 2007) f) invited to speak at the NSF workshop on bridging the gap between physical layer and networking layers g) invited Caltech talk, Nov. 2007

Faculty Name: P.Vijay Kumar


Spotlight: Designed a reduced delay physical layer fusion algorithm for a sensor network. Obtained a finite-step algorithm for a convex optimisation problem with linear ascending constraints. Established tight connections between guessing and compression with applications to secret-key cryptosystems.

Faculty Name: Rajesh Sundaresan


Spotlight: Working on wireless sensor networks and their usage in Social Network Analysis(SNA) and Residential layouts for safety and security reasons.

Faculty Name: S.V.Gopalaiah


Spotlight: Errors in Delay Differentiation in Statistical Multiplexing: The scheduling of packet transmissions from several packet queues in a statistical multiplexer with specified packet arrival processes, so that average packet delays in the queues approach specified feasible target delays, has been achieved through Markov-decision-theoretic policies; it has been shown in general that the error in average delay for at least one queue has a lower bound that varies inversely as the square root of the averaging duration

Faculty Name: Utpal Mukherji


Spotlight: Developed and analysed an energy efficient algorithm for detection of change via sensor networks. Modelling of TCP and computing its throughput via UMTS network. Providing QoS to TCP and real time connections via multihop wireless networks. Obtained new Joint source-channel coding schemes and analyzed some existing ones on a MAC.

Faculty Name: Vinod Sharma


Spotlight: Stochastic models for IEEE 802.11 Wireless LANs: generalised fixed point analysis, modeling of packet telephony, TCP transfers, and streaming video.

Faculty Name: Anurag Kumar


Spotlight: Developed an approach to design an ultra wideband conformal antenna using microstrip geometry

Faculty Name: K.J.Vinoy


Spotlight: Developed a systematic approach to design multi-frequency antennas using multiple concentric rings

Faculty Name: K.J.Vinoy


Spotlight: Designed algorithms for structuring interference in uplink CDMA. Designed an equaliser for DL OFDM systems.

Faculty Name: Rajesh Sundaresan


Spotlight: Opportunistic scheduling of wireless links. Modelling of Wireless fluid queues. Analysis and convergence of LMS Decision feedback equalizers.

Faculty Name: Vinod Sharma


Spotlight: Modeling of TCP: over lossy and fading wireless links; long range dependent traffic and closed loop behaviour of buffers

Faculty Name: Anurag Kumar


Spotlight: • Co-invented PILOT technology as PhD thesis for time and memory efficient circuit-electromagnetic simulation, 10-100x faster than existing commercial solutions. This technology was subsequently licensed by the University of Washington and forms the key technology behind Nimbic (formerly Physware) products.

Faculty Name: Dipanjan Gope


Spotlight: A new approach for miniaturization of bandpass filters on coplanar waveguides was developed. This uses spiral shaped slots on the ground traces

Faculty Name: K.J.Vinoy


Spotlight: Identified asymptotic performance of robust guessing strategies with potential applications to secret key cryptosystems.

Faculty Name: Rajesh Sundaresan


Spotlight: Analysis of LMS linear equalizer for fading channels where fading channel is modelled as an ARMA(2) process. Alogorithms for routing and centralized scheduling in IEEE 802.16 Mesh networks. Diversity-multiplexing trade-off for Wireless channels with space-time power control.

Faculty Name: Vinod Sharma


Spotlight: Developed an algorithm to measure fractal lacunarity for geometries with different fractal dimansions.

Faculty Name: K.J.Vinoy


Spotlight: Designed an HSDPA receiver with equaliser. Two patent applications made in 2006.

Faculty Name: Rajesh Sundaresan