Mohammad Alaee-Kerahroodi, Interdisciplinary Centre for Security, Reliability and Trust (SnT), University of Luxembourg.
Bhavani Shankar M. R., Interdisciplinary Centre for Security, Reliability and Trust (SnT), University of Luxembourg.
- M. Alaee-Kerahroodi, M. Modarres-Hashemi, M.M. Naghsh, “Designing Binary Sequence Sets for MIMO Radar Systems”, IEEE Transaction on Signal Processing, Volume : 67 , Issue : 13 , July1, 1 2019.
- Kumar Vijay Mishra, M. R. Bhavani Shankar, Visa Koivunen, Björn Ottersten and Sergiy A. Vorobyov, “Toward millimeter wave joint radar-communications: A signal processing perspective,” IEEE Signal Processing Magazine, vol. 36(5), pp. 100-114, 2019.
- Sayed Hossein Dokhanchi, M. R. Bhavani Shankar, Kumar Vijay Mishra and Björn Ottersten, “A mmWave automotive joint radar-communications system,” IEEE Transactions on Aerospace and Electronic Systems, vol. 55(3), pp. 1241-1260, 2019.
- M. Alaee-Kerahroodi, A. Aubry, A. De-Maio, M.M. Naghsh and M. Modarres-Hashemi, “A Coordinate Descent Framework to Design Low PSL/ISL Sequences”, IEEE Transactions on Signal Processing, Volume : 65 , Issue : 22, Nov.15, 2017.
- M. M. Naghsh, M. Modarres-Hashemi, M. Alaee-Kerahroodi, and E. H. M. Aian, “An information theoretic approach to robust constrained code design for MIMO radars,” IEEE Transactions on Signal Processing,vol. 65, Issue 14, pp. 3647 – 3661. Year 2017.
The main goal of the tutorial is to provide the audience with a bouquet of optimization techniques to address different challenging waveform design problems in classical and emerging Multiple Input Multiple Output (MIMO) radar systems, under practical constraints.
Waveform design plays a key role in enhancing classical radar tasks including target detection and parameter estimation. Further, waveform design is a key enabler of the emerging paradigm on joint radar-communications. Different applications warrant different performance metrics; this coupled with the advent of MIMO radar makes the waveform design more challenging. Particularly, in the emerging scenario of self-driving automotive applications, towards enhancing safety and comfort, high spatial resolution is achieved using the colocated MIMO virtual array by maintaining orthogonality between the transmit waveforms. Further, waveform diversity can also be used to obtain low-probability-of-intercept (LPI) radar properties. Nevertheless, the static use of a fixed waveform reduces efficiency due to limited or no adaptation to the dynamic environment as well as vulnerability to electronic attacks highlighting the need for multiple and diverse waveforms exhibiting specific features.
In this context, the tutorial focusses on key applications and highlights a variety of optimization approaches including coordinate descent (CD) and majorization minimization (MM), dealing with important applications in radar including 1) enhancing angular resolution using sets of orthogonal sequences, 2) SINR enhancement with joint design of space-time transmit and receive weights, 3) enabling a joint radar-communications paradigm through the transmit waveform design. To further bring the optimization closer to implementation and early adaptation in systems, practical constraints, such as finite energy, unimodularity (or being constant-modulus) and finite or discrete-phase alphabet are included in the optimization problem as constraints. The diversity of design metrics and signal constraints lays the groundwork for many interesting research projects in waveform optimization.
While several seminal works have been published, a few previous “IRS” tutorials have focused on the optimization algorithms dealing with the various applications of active sensing. After attending the tutorial, participants will be able to understand:
- An overview of relevant theoretical bases and algorithms from optimization theory considered in the state-of-the-art waveform design.
- Current challenges and design criteria associated with waveform design in classical and emerging radar systems.
- Key hardware constraints of the practical radar systems and their consideration in the optimization formulation.
- An insight into formulation of waveform design optimization problems in modern radar systems and a few approaches towards finding a solution.
We will present this tutorial in two slots and different parts as listed below:
- Slot 1 (1 hour and 40 minutes)
- Part I: A brief review of optimization principles, active sensing scenarios and problem formulation (50 mins): This part begins by describing and illustrating principles of convex and non-convex optimization theory. Next, we consider casting various design problems in active sensing systems. More precisely, we address several scenarios like PSL/ISL minimization for classical radar systems, designing sets of orthogonal sequences for emerging MIMO radar systems, joint sensing and communications and so on. In this context, emphasis on the objective functions and constraint sets of the associated problems.
- Part II: CD optimization framework for transceiver design (50 mins): The CD based methods are intuitively appealing and simple to implement, yet they have shown powerful performance in emerging large-scale signal processing, machine-learning, regression, compressed sensing, and radar applications. The idea behind CD is not to tackle the original problem directly, but by iteratively optimizing it over a single coordinate, while keeping the other coordinates fixed. The most important advantage of the CD method is that the minimization of a multi-variable function can be achieved minimizing it along one direction at a time, i.e., solving a set of potentially simpler uni-variate sub-problems in a loop. Using this framework, we illustrate how to apply CD method on the design problem introduced in the previous part.
- Slot 2 (1 hour and 40 minutes)
- Part III: Waveform optimization in mm-Wave sensing and communications (40 mins): In this part we introduce the driving factors for mm-Wave spectrum sharing, low-cost design and differences with respect to cm-Wave joint sensing-communications. The need for synergetic waveform design accomplishing radar and communication tasks will be highlighted. Focussing on the automotive scenario, different topologies and related challenges on waveform design will be presented. Waveform design based on aforementioned methodologies will be presented and the gains achieved will be discussed.
- Part IV: MM optimization framework for waveform design (40 mins): The MM based methods introduced for various transceiver design problems in active sensing systems will be presented in this part. The idea is to address a difficult optimization problem indirectly, by finding a surrogate function that makes the optimization problem “easy” (or, in any case, easier than the directly solving the original problem). We illustrate tricks for finding surrogate functions and the key aspects in this framework through a variety number of examples.
- Part V: Summary and open challenges (20 mins): The aforementioned optimization methodologies have gained growing popularity in various applications. Some of these will be mentioned, and a summary of the introduced methods as well as the remaining challenges will be discussed in this part.