Jayanta Roy

1. Surveys for pulsars and transients.
Studies of pulsars yield a better understanding of a variety of physics problems, from acceleration of particles in ultra-strong magnetic fields (primarily via studies of emission properties of normal pulsars, having spin period > 30 ms) to probes of ultra-dense matter (mostly via studying the timing properties of millisecond pulsars, having spin period < 30 ms, that are very stable rotators). This stability, a compactness second only to black holes, and the fact that millisecond pulsars (MSPs) are often in binary systems, makes them ideal laboratories to test the physics of gravity and also as detectors for long-wavelength gravitational waves. We have been conducting both targeted and blind surveys for pulsars with the GMRT, resulting in the discovery of 20 pulsars (including 9 MSPs) till date. The GMRT High Resolution Southern Sky (GHRSS) survey is an off-Galactic-plane (|b| >5) survey in the declination range −40 deg to −54 deg at 322 MHz. With 60% of the survey completed we have discovered 13 pulsars, one of which is a MSP. The GHRSS survey data are concurrently being processed by a frequency-domain acceleration search algorithm using CPUs and by a time-domain acceleration search algorithm using GPUs, thus providing a detailed comparison between the algorithms, which can be a vital input to the SKA choices. From the targeted survey of unassociated Fermi gamma-ray sources with the GMRT, we discovered 7 MSPs including the first Galactic field MSPs discovered at the GMRT. A coherently dedispersed multi-pixel coherent search for pulsars in Globular clusters is also being carried out. Fast Radio Bursts (FRBs) are observed at the locations of energetic, and possibly cataclysmic events, making them useful probes of extreme states of gravity, pressure, temperature and magnetic fields. Magnetar flares, super-giant pulses from neutron stars and pulsar-planet systems are some of the models for the extra-galactic origin of fast radio bursts, at cosmological distances. Besides the Fourier-based periodicity searches, single pulse searches are performed for these survey data to detect FRBs. The full GHRSS survey is expected to detect ∼ 4 FRBs at fluence of 3 Jy-ms. Identification of such FRBs with host galaxies through simultaneous localisation using the GMRT interferometer can provide vital inputs in understanding the nature of FRBs.
2. Timing of pulsars.
The diverse and unique properties of the newly discovered pulsars can only be revealed through follow up timing studies. We are conducting regular timing studies of the 20 pulsars discovered by us with the GMRT. Some of these systems can be timed sufficiently precisely to aid the International Pulsar Timing Array (IPTA) sensitivity to detection of Gravitational Waves (GWs). We are also working towards a dual-frequency coherently-dedispersed timing approach to improve the sensitivity of NANOGrav (the North American Nanohertz Observatory for Gravitational Waves) to a stochastic GW background, aiming to benefit by a better modeling of the dispersive interstellar medium effects to improve the timing precision.
3. Astrometry of pulsars.
Motivated by the need for rapid astrometry of newly discovered pulsars, we have developed a coherently-dedispersed gating correlator at the GMRT. Using this gating correlator, unambiguous localisation of newly discovered pulsars (even faint MSPs) can be achieved with arcsecond accuracy in the on-off gated image plane. We have localised 11 new pulsars in the gated image plane within ± 10”. Immediate knowledge of such precise positions enables the use of sensitive coherent beams of array telescopes for follow-up timing observations, while allowing timing-independent rapid astrometric measurements to break the degeneracy caused by covariance between pulsar position and rotational, binary parameters.
4. Studies of transitional millisecond pulsars.
Low-mass X-ray binaries (LMXBs) and radio millisecond pulsars (MSPs) are linked through stellar and binary evolution, where during the mass accretion, phase these systems are bright X-ray sources and when mass transfer eventually declines, a pulsar powered by rotation of its magnetic field turns on. PSR J1227-4853 (discovered by us with the GMRT) is only the third system showing direct evidence for a transition from an LMXB to an MSP. Following the discovery of radio millisecond pulsation, radio timing observations allowed precise determination of the rotational and orbital parameters of this system. PSR J1227-4853 is one of the few eclipsing binary systems, where simultaneous imaging and time-domain studies were done to directly probe the cause of eclipse. Using this long-term radio timing follow-up, we have also detected gamma-ray pulsations after the transition, using data from the Large Area Telescope (LAT) on board the Fermi Gamma-ray Space Telescope. PSR J1227-4853 may exhibit reverse transitions resulting in the disappearance of radio pulsations. The regular radio timing study with the GMRT and study of gamma-ray flux light curves exhibits that the pulsar is still in the active-MSP state, and also reveals higher order orbital perturbations and significant proper motion.
5. Pulsars and fast transients survey instrument for the Square Kilometre Array.
Pulsar survey(s) with the Square Kilometre Array (SKA), with its unprecedented sensitivity and high figure of merit, will result in the discovery of a large population of pulsars, including pulsars of previously unknown types. Equipped with a wide field-of-view, an unprecedented sensitivity, and a multi-beaming capability, coupled with real-time searches for pulsars and transients, SKA is expected to bring a renaissance in the field of time-domain processing. It has been predicted that SKA will reveal around 20000 of pulsars including ~6000 millisecond pulsars, aiding the sensitivity of pulsar-timing searches for gravitational waves by orders of magnitude. It is also expected that time-domain processing with the SKA will detect a large number of fast radio bursts, providing a better understanding of the population and their progenitors. As a member of the SKA Time Domain Team, I am involved in the design of the Pulsar Search Sub-element (PSS) of the SKA Central Signal Processing (CSP), which is aimed to process 1500 beams for SKA-Mid array and 500 beams for SKA-Low array, in real-time, searching for pulsars and fast transients. The proposed SKA non-imaging cluster will consist of around 1000 processing nodes, each hosting multiple compute accelerators like state-of-the-art GPUs and FPGAs, needing to provide 10 PetaOps of compute power, while crunching a data throughput of 60 PetaBytes per day. Energy-efficient computing is one of the main design aims for the PSS prototyping efforts, which will make the PSS five times greener than the current greenest supercomputer in the world. Finally, the PSS system fully based on COTS, will be equipped to get real-time detections of a wide variety of pulsars and fast radio bursts, including interesting individual systems like pulsar/black-hole binaries, pulsars orbiting the super-massive black hole at the centre of the Milky Way, sub-millisecond pulsars, etc.
6. Radio Astronomy Signal Processing using High-Performance Computing.
Recent advancements in computing power, enhanced memory bandwidth, high speed networking and increased throughput of storage media have revolutionized the high performance computing associated with radio astronomy. The computing requirements of radio astronomy signal processing, for a multi-element interferometric array, are very well suited for implementation on multi-processor vectorised machines. We designed and implemented a 32-antenna, 33 MHz, dual polarization, fully real-time software back-end (GSB) for the GMRT. I am also involved with the design of a pulsar beamformer for the eMerlin array. Faster development cycles, more flexibility and easy upgrade of the software-based design stand to gain more from Moore's law. Software-based processing also has a growing importance for the next generation radio telescopes with a far larger number of elements and bandwidth, due to the massive increase in computing power aided by accelerators like GPUs and FPGAs.


Document Actions