RESEARCH INTERESTS

5G systems are more than just the next step in the evolution of mobile communications. As an enabler of the networked society, 5G networks need to provide capacity and capabilities not only for voice and data communication, but also for new use cases, device-to-device communications and applications for a connected society. Basically, 5G will play a crucial role in the operation of the society. Therefore, security and privacy are essential for the success of 5G. 5G is likely to generate new requirements because of new business and trust models, new service delivery models, a larger attack surface and an increased concern for privacy. IoT security will also play an important role in 5G security. We are looking at the security of IoT systems from the perspective of 5G.

Physical layer security is becoming an emerging technique to complement and significantly improve the communication security of wireless networks. Compared to traditional cryptographic approaches, physical layer security is a fundamentally different paradigm where secrecy is achieved by exploiting the physical layer properties of the communication system. Current research interest lies in the computation of the secrecy outage probabilities for different relay systems in the presence of an eavesdropper, with and without the knowledge of eavesdropper's channel state information (CSI). We are also looking at the sum secrecy rate maximization for 1) subcarrier usage policy at the source and the relay and 2) power allocation among subcarriers under a total various power constraints. Optimal relay selection is another interesting problem we are looking at.

What is the entropy of entropy? Or, in layman's language: what is the randomness of randomness? We call it 'Superinformation!'. We use Superinformation to segregate introns from exons (in the world of genomic studies) and predict the next economic crisis (in the world of econophysics)!

We are looking at game theoretic approached to 'predict' attacks on cyber physical systems and wireless sensor networks. We analyze what parameters are of interest, and how should we change our strategy to outwit the adversary?

The current research involves various aspects of UWB communications including channel sounding, channel modeling, transceiver design for specific channel models and Multi Input Multi Output (MIMO) systems for UWB. For the first time, channel sounding and measurement facility has been set up in IIT Delhi. We are also exploring the design of optimal Space Time Trellis Codes (STTC) for UWB communications. Another problem that we are investigating is the design of Smart Antennas for UWB. A DST project is currently underway which deals with the Design and development of a UWB Communication system. Visit the Homepage of the Ultra Wide Band (UWB) Group

We are investigating a novel method for generating trellis codes based on recursive nonlinear equations. Using this method, trellis codes having good code rates (R = k/n) can be constructed. At the same time, the method allows the code-designer to construct a trellis with a large free distance, dfree, at the cost of a larger number of states in the trellis. It is shown that for code rate, R, less than a critical rate, Rc, the free distance can be made as large as desired for large n. This design methodology provides the freedom to play with the code performance and the code rate of the trellis code.

We are working at new algorithms to determine the optimal locations of base stations, given the coverage by individual base stations. One approach has been to use the concepts of dynamic programming. Using this algorithm, a 20-25% decrease in the number of base stations required has been observed for simulated environments. Yet another approach is to use Genetic Algorithms. A software tool has been developed which assists in optimal base station placement.

An entirely new approach to interference reduction is being investigated using Trellis Coded Modulation (TCM). Traditionally, TCM has been used for error control. Our technique puts TCM to an entirely different application of co-channel interference reduction. A software tool has been developed which determines the interference level given the system architecture and calculates the reduced interference levels upon deployment of TCM schemes.

We are investigating the unpredictability of chaos functions to generate keys for cryptography. This is different from generating keys using a pseudo random number generator, which is predictable. It was found that the level of security can be increased dramatically using a set of chaos functions. We are also looking at Public Key Encryption using chaos functions. Here is a list of Patents related to this work.

LMDS is an upcoming broadband wireless access standard operating at 28 GHz and 42 GHz. We have carried out a systematic study of the co-channel interference resulting due to frequency reuse was carried out for the LMDS architecture. We have also proposed several techniques for interference mitigation. One of the technique uses the reorientation of the receiver antennas to alternate base station under system availability constraint.

The research involves the design space exploration for optimizing both memory and power for Turbo Encoder and Decoder, Wireless Sensor Networks and Encoder and Decoder for LDPC Codes, Smart low-power antenna (now we call them Green Antennas!). What is the best way to transmit data for 5G systems if you wish to account for the processor power?