Search

Edge-AI computing requires high energy efficiency, low power consumption, and relatively high flexibility and compact area, challenging the AI-chip design. This work presents a 0.96 pJ/SOP heterogeneous neuromorphic system-on-chip (SoC) with fullerene-like interconnection topology for edge-AI computing. The neuromorphic core integrates different technologies to augment computing energy efficiency, including sparse computing, partial membrane potential updates, and non-uniform weight quantization. Multiple neuromorphic cores and multi-mode routers form a fullerene-like network-on-chip (NoC). The average degree of communication nodes exceeds traditional topologies by 32%, with a minimal degree variance of 0.93, allowing advanced decentralized on-chip communication. Additionally, the NoC can be scaled up through extended off-chip high-level router nodes. A RISC-V CPU and a neuromorphic processor are tightly coupled and fabricated within a 5.42 mm^2 die area under 55 nm CMOS technology. The chip has a low power density of 0.52 mW/mm^2, reducing 67.5% compared to related works, and achieves a high neuron density of 30.23 K/mm^2. Eventually, the chip is demonstrated to be effective on different datasets and achieves 0.96 pJ/SOP energy efficiency., Comment: 5 pages, 8 figures

We present the discovery and timing results of four pulsars discovered in a pilot survey at intermediate Galactic latitudes with the Five-hundred Aperture Spherical Telescope (FAST). Among these pulsars, two belong to the category of millisecond pulsars (MSPs) with spin periods of less than 20 ms. The other two fall under the classification of "mildly recycled" pulsars, with massive white dwarfs as companions. Remarkably, this small survey, covering an area of 4.7 $deg^2$ , led to the discovery of four recycled pulsars. Such success underscores the immense potential of future surveys at intermediate Galactic latitudes. In order to assess the potential yield of MSPs, we conducted population simulations and found that both FAST and Parkes new phased array feed surveys, focusing on intermediate Galactic latitudes, have the capacity to uncover several hundred new MSPs., Comment: 7 pages, 4 figures, 2 tables, accepted to ApJ

Pulsars radiate radio signals when they rotate. However, some old pulsars often stop radiating for some periods. The underlying mechanism remains unknown, while the magnetosphere during nulling phases is hard to probe due to the absence of emission measurement. Here we report the detection and accurate polarization measurements of sporadic weak narrow dwarf pulses detected in the ordinary nulling state of pulsar B2111+46 via the Five-Hundred-Meter Aperture Spherical radio Telescope (FAST). Further analysis shows that their polarization angles follow the average polarization angle curve of normal pulses, suggesting no change of magnetic field structure in the emission region in the two emission states. Whereas radio emission of normal individual pulses is radiated by a thunderstorm of particles produced by copious discharges in regularly formed gaps, dwarf pulses are produced by one or a few raindrops of particles generated by pair production in a fragile gap of this near-death pulsar., Comment: 37 pages, 15 figures. Authors version. Great FAST pulsar data. Share full-text access: https://rdcu.be/djLSL

Analysis of helicon wave heating and current drive capability in EXL-50 spherical torus plasmas has been conducted. It is found that the driven current increases with the launched parallel refractive index $n_{||}$ and peaks around $n_{||} = 4.0$ when the frequency of the helicon wave is between 300~MHz and 380~MHz. The helicon wave current drive efficiency shows a relatively stable upward trend with increasing plasma temperature. Moreover, the driven current decreases as the plasma density increases. We also analyzed the current drive with helicon waves of 150~MHz and 170~MHz and found that the driven current at a lower frequency was lower than that at a higher frequency. A positive proportional relationship exists between the driven current and $n_{||}$. Besides, as $n_{||}$ increases, the profile of the driven current becomes wider. Finally, the effect of the scrape-off layer (SOL) region on the helicon wave current drive was also investigated.

We present the first measurement of pulse scattering close to the eclipse region of PSR B1957+20, which is in a compact binary system with a low-mass star. We measured pulse scattering time-scales up to 0.2 ms close to the eclipse and showed that it scales with the dispersion measure (DM) excess roughly as $\tau\propto\Delta{\rm DM}^{2}$. Our observations provide the first evidence of strong scattering due to multi-path propagation effects in the eclipsing material. We show that Kolmogorov turbulence in the eclipsing material with an inner scale of $\sim100$ m and an outer scale of the size of the eclipse region can naturally explain the observation. Our results show that the eclipsing material in such systems can be highly turbulent and suggest that scattering is one of the main eclipsing mechanisms at around 1.4 GHz., Comment: 8 pages, 4 figures, MNRAS accepted

Based on the concepts of quantum synchronization and quantum phase synchronization proposed by A. Mari \textit{et al.} in Phys. Rev. Lett. 111, 103605 (2013), we introduce and characterize the measure of a more generalized quantum synchronization called quantum $\varphi$-synchronization under which the pairs of variables have the same amplitude and possess the same $\varphi$ phase shift. Naturally, quantum synchronization and quantum anti-synchronization become special cases of quantum $\varphi$-synchronization. Their relations and differences are also discussed. To illustrate these theories, we investigate the quantum $\varphi$-synchronization and quantum phase synchronization phenomena of two coupled optomechanical systems with periodic modulation and show that quantum $\varphi$-synchronization is more general as a measure of synchronization. We also show the phenomenon of quantum anti-synchronization when $\varphi=\pi$., Comment: 6 pages, 5 figures, comments are wellcome

Zinc antimony stands out among thermoelectrics because of its very low lattice thermal conductivity, close to the amorphous limit. Understanding the physical reason behind such an unusual crystal property is of fundamental interest for the design of new thermoelectric materials. In this work we report the results of atomistic computer simulations on experimentally determined beta-Zn4Sb3 structures. We find a remarkably anharmonic behavior of Zn atoms that could be responsible for the low thermal conductivity of Zn4Sb3: their movement, better explained as diffusive, does not contribute to thermal conduction. Moreover, phonon transport is impeded by a lack of coupling between Zn and Sb atoms in crystalline positions., Comment: 11 pages, 3 figures

The emission geometries, e.g. the emission region height, the beam shape, and radius-to-frequency mapping, are important predictions of pulsar radiation model. The multi-band radio observations carry such valuable information. In this paper, we study two bright pulsars, (PSRs B0329+54 and B1642-03) and observe them in high frequency (2.5 GHz, 5 GHz, and 8 GHz). The newly acquired data together with historical archive provide an atlas of multi-frequency profiles spanning from 100 MHz to 10 GHz. We study the frequency evolution of pulse profiles and the radiation regions with the these data. We firstly fit the pulse profiles with Gaussian functions to determine the phase of each component, and then calculate the radiation altitudes of different emission components and the radiation regions. We find that the inverse Compton scattering (ICS) model can reproduce the radiation geometry of these two pulsars. But for PSR B0329+54 the radiation can be generated in either annular gap (AG) or core gap (CG), while the radiation of PSR B1642-03 can only be generated in the CG. This difference is caused by the inclination angle and the impact angle of these two pulsars. The relation of beaming angle (the angle between the radiation direction and the magnetic axis) and the radiation altitudes versus frequency is also presented by modelling the beam-frequency evolution in the ICS model. The multi-band pulse profiles of these two pulsars can be described well by the ICS model combined with the CG and AG.

The very small braking index of PSR J1734-3333, $n=0.9\pm0.2$, challenges the current theories of braking mechanisms in pulsars. We present a possible interpretation that this pulsar is surrounded by a fall-back disk and braked by it. A modified braking torque is proposed based on the competition between the magnetic energy density of a pulsar and the kinetic energy density of a fall-back disk. With this torque, a self-similar disk can fit all the observed parameters of PSR J1734-3333 with natural initial parameters. In this regime, the star will evolve to the region having anomalous X-ray pulsars and soft gamma repeaters in the $P-\dot{P}$ diagram in about 20000 years and stay there for a very long time. The mass of the disk around PSR J1734-3333 in our model is about $10M_{\oplus}$, similar to the observed mass of the disk around AXP 4U 0142+61., Comment: 8 pages, 3 figures, 1 table

Timing observations from the Parkes 64-m radio telescope for 165 pulsars between 1990 and 2011 have been searched for period glitches. A total of 107 glitches were identified in 36 pulsars, where 61 have previously been reported and 46 are new discoveries. Glitch parameters were measured by fitting the timing residual data. Observed relative glitch sizes \Delta\nu_g/\nu range between 10^-10 and 10^-5, where \nu = 1/P is the pulse frequency. We confirm that the distribution of \Delta\nu_g/\nu is bimodal with peaks at approximately 10^-9 and 10^-6. Glitches are mostly observed in pulsars with characteristic ages between 10^3 and 10^5 years, with large glitches mostly occurring in the younger pulsars. Exponential post-glitch recoveries were observed for 27 large glitches in 18 pulsars. The fraction Q of the glitch that recovers exponentially also has a bimodal distribution. Large glitches generally have low Q, typically a few per cent, but large Q values are observed in both large and small glitches. Observed time constants for exponential recoveries ranged between 10 and 300 days with some tendency for longer timescales in older pulsars. Shorter timescale recoveries may exist but were not revealed by our data which typically have observation intervals of 2 - 4 weeks. For most of the 36 pulsars with observed glitches, there is a persistent linear increase in \dot\nu in the inter-glitch interval. Where an exponential recovery is also observed, the effects of this are superimposed on the linear increase in \dot\nu. In some cases, the slope of the linear recovery changes at the time of a glitch. The \ddot\nu values characterising the linear changes in \dot\nu are almost always positive and, after subtracting the magnetospheric component of the braking, are approximately proportional to the ratio of |\dot\nu| and the inter-glitch interval, as predicted by vortex-creep models., Comment: Accepted for publication in MNRAS

The origin of the 6.67 hr period X-ray source, 1E161348-5055, in the young supernova remnant RCW 103 is puzzling. We propose that it may be the descendant of a Thorne-Zytkow Object (TZO). A TZO may at its formation have a rapidly spinning neutron star as a core, and a slowly rotating envelope. We found that the core could be braked quickly to an extremely long spin period by the coupling between its magnetic field and the envelope, and that the envelope could be disrupted by some powerful bursts or exhausted via stellar wind. If the envelope is disrupted after the core has spun down, the core will become an extremely long-period compact object, with a slow proper motion speed, surrounded by a supernova-remnant-like shell. These features all agree with the observations of 1E161348-5055. TZOs are expected to have produced extraordinary high abundances of lithium and rapid proton process elements that would remain in the remnants and could be used to test this scenario., Comment: 6 pages

(adapted)Considering recent observations challenging the traditional magnetar model, we explore the wind braking of magnetars. There is evidence for strong multipole magnetic fields in active magnetars, but the dipole field inferred from spin down measurements may be strongly biased by a particle wind. Recent challenging observations of magnetars may be explained naturally in the wind braking scenario: (1) The supernova energies of magnetars are of normal value; (2) The non-detection in Fermi observations of magnetars; (3) The problem posed by the low-magnetic field soft gamma-ray repeaters; (4) The relation between magnetars and high magnetic field pulsars; (5) A decreasing period derivative during magnetar outbursts. Transient magnetars may still be magnetic dipole braking. This may explain why low luminosity magnetars are more likely to have radio emissions. In the wind braking scenario, magnetars are neutron stars with strong multipole field. For some sources, a strong dipole field may be no longer needed. A magnetism-powered pulsar wind nebula and a braking index smaller than three are the two predictions of the wind braking model., Comment: 34 pages, 1 table, 6 figures, accepted by ApJ

To investigate the missing compact star of Supernova 1987A, we analyzed both the cooling and the heating processes of a possible compact star based on the upper limit of observational X-ray luminosity. From the cooling process we found that a solid quark-cluster star, which has a stiffer equation of state than that of conventional liquid quark star, has a heat capacity much smaller than a neutron star. It can cool down quickly, which can naturally explain the non-detection of a point source (neutron star or quark star) in X-ray band. On the other hand, we consider the heating process from magnetospheric activity and possible accretion, and obtain some constraints to the parameters of a possible pulsar. We conclude that a solid quark-cluster star can be fine with the observational limit in a large and acceptable parameter space. A pulsar with a short period and a strong magnetic field (or with a long period and a weak field) would has luminosity higher than the luminosity limit if the optical depth is not large enough to hide the compact star. The constraints of the pulsar parameters can be tested if the central compact object in 1987A is discovered by advanced facilities in the future., Comment: 5 pages, 2 figures (submitted to MNRAS)

The relationship between pulsar-like compact stars and gravitational waves is briefly reviewed. Due to regular spins, pulsars could be useful tools for us to detect ~nano-Hz low-frequency gravitational waves by pulsar-timing array technique; besides, they would also be ~kilo-Hz high-frequency gravitational wave radiators because of their compactness. The wave strain of an isolate pulsar depends on the equation state of cold matter at supra-nuclear densities. Therefore, a real detection of gravitational wave should be very meaningful in gravity physics, micro-theory of elementary strong interaction, and astronomy., Comment: 11 pages, 3 figures; in: Gravitation and Astrophysics (Proceedings of the IX Asia-Pacific International Conference, 29 June - 2 July, 2009, Wuhan), eds. J. Luo, Z. B. Zhou, H. C. Yeh, and J. P. Hsu, World Scientific, p.162-172

The inverse Compton scattering (ICS) model can explain various pulse profile shapes and diversity of pulse profile evolution based on the mechanism that the radio emission is generated through inverse Compton scattering between secondary relativistic particles and radio waves from polar gap avalanches. In this paper, we study the parameter space of ICS model for 15 pulsars, which share the common pulse profile evolution phenomena that the pulse profiles are narrower at higher observing frequencies. Two key parameters, the initial Lorentz factor and the energy loss factor of secondary particles are constrained using the least square fitting method, where we fit the theoretical curve of the "beam-frequency mapping" of the ICS model to the observed pulse widths at multiple frequencies. The uncertainty of the inclination and viewing angles are taken into account in the fitting process. It is found that the initial Lorentz factor is larger than 4000, and the energy loss factor is between 20 and 560. The Lorentz factor is consistent with the prediction of the inner vacuum gap model. Such high energy loss factors suggest significant energy loss for secondary particles at altitudes of a few tens to hundreds of kilometers., Comment: 32 pages,18 figures, Accepted by the ApJ

The Vela pulsar represents a distinct group of {\gamma}-ray pulsars. Fermi {\gamma}-ray observations reveal that it has two sharp peaks (P1 and P2) in the light curve with a phase separation of 0.42 and a third peak (P3) in the bridge. The location and intensity of P3 are energy-dependent. We use the 3D magnetospheric model for the annular gap and core gap to simulate the {\gamma}-ray light curves, phase-averaged and phase-resolved spectra. We found that the acceleration electric field along a field line in the annular gap region decreases with heights. The emission at high energy GeV band is originated from the curvature radiation of accelerated primary particles, while the synchrotron radiation from secondary particles have some contributions to low energy {\gamma}-ray band (0.1 - 0.3 GeV). The {\gamma}-ray light curve peaks P1 and P2 are generated in the annular gap region near the altitude of null charge surface, whereas P3 and the bridge emission is generated in the core gap region. The intensity and location of P3 at different energy bands depend on the emission altitudes. The radio emission from the Vela pulsar should be generated in a high-altitude narrow regions of the annular gap, which leads to a radio phase lag of ~ 0.13 prior to the first {\gamma}-ray peak., Comment: 9 pages, 8 figures, 2 tables, accepted by ApJ

Anomalous X-ray Pulsars and Soft Gamma-Ray Repeaters have been generally recognized as neutron stars with super strong magnetic fields, namely "magnetars". The "magnetars" manifest that the luminosity in X-ray band are larger than the rotational energy loss rate, i.e. $L_{X}>\dot {E}_{\rm rot}$, and then the radiation energy is coming from the energy of magnetic field. Here it is argued that magnetars may not really exist. Some X-ray and radio observational results are contradicted with the magnetar model. (1) The X-ray luminosity of PSR J1852+0040 is much larger than the rotational energy loss rate ($L_{X}/\dot {E}_{\rm rot}\simeq 18)$, but the magnetic field is just $3.1\times 10^{11}$ G. Does this X-ray radiation energy come from the magnetic field? (2) In contrast to the above, the magnetic fields of radio pulsars J1847-0130 and PSR J1718-3718 are higher than that of AXP 1E 2259+586, why is the radiation energy of those two radio pulsars still coming from rotational energy? Furthermore, the magnetic field of the newly discovered SGR 0418+5729 with the lowest magnetic field is 3.0 e13 G, lower than the critical magnetic field $B_{\rm C}=4.414$ e13 G) (Esposito et al. 2010). (3) Some "magnetars" also emit normal transient radio pulses, what is the essential difference between radio pulsars and the "magnetars"? The observational fact arguments will be presented at first, then we discuss in what situation the conventional method to obtain magnetic field could not be correct.

Pulsed high energy radiation from pulsars is not yet completely understood. In this paper, we use the 3D self-consistent annular gap model to study light curves for both young and millisecond pulsars observed by the Fermi Gamma-ray Space Telescope. The annular gap can generate high energy emission for short-period pulsars. The annular gap regions are so large that they have enough electric potential drop to accelerate charged particles to produce gamma-ray photons. For young pulsars, the emission region is from the neutron star surface to about half of the light cylinder radius, and the peak emissivity is in the vicinity of the null charge surface. The emission region for the millisecond pulsars is located much lower than that of the young pulsars. The higher energy gamma-ray emission comes from higher altitudes in the magnetosphere. We present the simulated light curves for three young pulsars (the Crab, the Vela, the Geminga) and three millisecond pulsars (PSR J0030+0451, PSR J0218+4232, PSR J0437-3715) using the annular gap model. Our simulations can reproduce the main properties of observed light curves., Comment: 9 pages, 3 figures and 1 table; accepted for publication in MNRAS

Determining radiation location observationally plays a very important role in testing the pulsar radiation models. One-photon pair production in the strong magnetic field, $\gamma-e^{+}e^{1}$, is one of the important physical processes in pulsar radiation mechanisms. Photons near pulsar surface with sufficient energy will be absorbed in the magnetosphere and the absorption optical depth for these GeV $\gamma$-ray photons is usually large. In this paper, we include the aberrational, rotational and general relativistic effects and calculate the $\gamma$-B optical depth for $\gamma$-ray photons. Then we use the derived optical depth to determine the radiation altitude lower bounds for photons with given energies. As a case study, we calculate the lower bounds of radiation altitudes of Crab pulsar for photons with energy from 5 GeV to 1 TeV., Comment: 11 pages, 5 figures,Accepted by MNRAS

Pulsars have been recognized as normal neutron stars, but sometimes argued as quark stars. {\it Sub-millisecond pulsars, if detected, would play an essential and important role in distinguishing quark stars from neutron stars.} We focus on the formation of such sub-millisecond pulsars in this paper. A new approach to form a sub-millisecond pulsar (quark star) via accretion induced collapse (AIC) of a white dwarf is investigated here. Under this AIC process, we found that: (1) almost all the newborn quark stars could have an initial spin period of $\sim 0.1$ ms; (2) the nascent quark stars (even with a low mass) have sufficiently high spin-down luminosity and satisfy the conditions for pair production and sparking process to be as sub-millisecond radio pulsars; (3) in most cases, the timescales of newborn quark stars in the phase of spin period $< 1$ (or $<0.5$) ms can be long enough to be detected. As a comparison, an accretion spin-up process (for both neutron and quark stars) is also investigated. It is found that, quark stars formed through AIC process can have shorter periods ($\leq$ 0.5 ms); while the periods of neutron stars formed in accretion spin-up process must be longer than 0.5ms. Thus if a pulsar with a period less than 0.5 ms can be identified in the future, it should be a quark star., Comment: 10 pages, 4 figures, 4 tables, accepted by MNRAS

Pulsars have been recognized as normal neutron stars or quark stars. Sub-millisecond pulsars, if detected, would play an essential and important role in distinguishing quark stars from neutron stars. A key question is how sub-millisecond pulsars could form. Both sub-Keplerian (for neutron and quark stars) and super-Keplerian cases (only for quark stars, which are bound additionally by strong interaction) have been discussed in this paper in order to investigate possible ways of forming sub-millisecond pulsars. In the sub-Keplerian case, the equilibrium periods of both neutron and quark stars could be as low as ~0.4 ms when they are spun up through accretion in binary systems. In the super-Keplerian case, pulsars could very likely have an initial period of ~0.1 ms if quark stars with different mass could be formed from accretion-induced collapse (AIC) of white dwarfs. The timescale for a sub-millisecond pulsar to keep its period <1 ms is restricted due to gravitational wave radiation. We found that the timescales of neutron stars within sub-millisecond periods are approximately ~10 yr, but the timescales estimated (>10 Myr) for low-mass quark stars could be long enough for us to detect., Comment: 12pages, 1 figure, 3 table, this paper has been withdrawn

Many models for the pulsar radio and $\gamma$-ray emissions have been developed. The tests for these models using observational data are very important. Tests for the pulsar radio emission models using frequency-altitude relation are presented in this paper. In the radio band, the mean pulse profiles evolve with observing frequencies. There are various styles of pulsar profile - frequency evolutions (which we call as "beam evolution" figure), e.g. some pulsars show that mean pulse profiles are wider and core emission is higher at higher frequencies than that at lower frequencies, but some other pulsars show completely the contrary results. We show that all these "beam evolution" figures can be understood by the Inverse Compton Scattering(ICS) model (see Qiao at al.2001 also). An important observing test is that, for a certain observing frequency different emission components are radiated from the different heights. For the $\gamma$-ray pulsars, the geometrical method (Wang et al. 2006) can be used to diagnose the radiation location for the $\gamma$-ray radiation. As an example, Wang et al. (2006) constrain the $\gamma$-ray radiation location of PSR B1055-52 to be the place near the null charge surface. Here we show that Wang's result matches the proposed radiation locations by the annular gap model as well as the outer gap models., Comment: 6 pages, 5 figures, Proceedings of "Astrophysics of Compact Objects", 1th-7th, July, 2007

If the binding energy of the pulsar's surface is not so high (the case of a neutron star), both the negative and positive charges will flow out freely from the surface of the star. The annular free flow model for $\gamma$-ray emission of pulsars is suggested in this paper. It is emphasized that: (1). Two kinds of acceleration regions (annular and core) need to be taken into account. The annular acceleration region is defined by the magnetic field lines that cross the null charge surface within the light cylinder. (2). If the potential drop in the annular region of a pulsar is high enough (normally the cases of young pulsars), charges in both the annular and the core regions could be accelerated and produce primary gamma-rays. Secondary pairs are generated in both regions and stream outwards to power the broadband radiations. (3). The potential drop in the annular region grows more rapidly than that in the core region. The annular acceleration process is a key point to produce wide emission beams as observed. (4). The advantages of both the polar cap and outer gap models are retained in this model. The geometric properties of the $\gamma$-ray emission from the annular flow is analogous to that presented in a previous work by Qiao et al., which match the observations well. (5). Since charges with different signs leave the pulsar through the annular and the core regions, respectively, the current closure problem can be partially solved., Comment: 11 pages 2 figures, accepted by Chinese Journal of Astronomy and Astrophysics

The consistence of the distributions of pulsar's kick velocities from the model of GRB and from the pulsar observations is tested based on the supernova-GRB ($\gamma$-ray burst) association and under the assumption that the GRB asymmetric explosions produce pulsars. The deduced distribution of kick velocity from the model of GRB and the observed kick distribution of radio pulsars are checked by K-S test. These two distributions are found to come from a same parent population. This result may indicate that GRBs could be really related to supernova, and that the asymmetry of GRB associated with supernova would cause the kick of pulsars., Comment: 4 pages, 1 figure, accepeted by A&A

The main idea and procedure are presented for a new synthetical method on constraining 3-D structure of emission regions of pulsars. With this method the emission regions can be synthetically constrained by fitting multi-wavelength observations features, e.g. pulse widths, phase offsets between radio pulse profiles and high energy light curves, and radio polarization properties. The main technique is based on numerically calculating the emission directions along any open field line via including aberration and retardation effects under a certain magnetic field configuration, e.g. vacuum dipole field. It can be easily extended by involving some further effects or magnetic field line configuration. The role of observed linear polarization properties in the method is discussed. The future application to known X/gamma-ray and radio pulsars will be helpful to test and improve current theoretical models on pulsar emission., Comment: Proceedings of the 363. WE-Heraeus Seminar on: Neutron Stars and Pulsars (Posters and contributed talks) Physikzentrum Bad Honnef, Germany, May.14-19, 2006, eds. W.Becker, H.H.Huang, MPE Report 291, pp.134-136

The pulse profile of pulsar gives geometric information about pulsar's radiation model. After investigating the pulse profiles of PSR B1642-03 and PSR B0950+08, we calculate the ratios of beam width and the emission height between different frequencies. We find that the ratios are almost constants as inclination angle $\alpha$ changes from $0^{\circ}$ to $90^{\circ}$. The ratios can be used to test pulsar's radiation model. In particular it can well constrain the parameter space of the inverse Compton scattering (ICS) model. These constrained parameters indicate some physical implication for the ICS model., Comment: 7 pages, 6 figures, submitted to "Advances in Space Research" (Proceedings of COSPAR 2006)

The large-scale magnetic field of our Galaxy can be probed in three dimensions using Faraday rotation of pulsar signals. We report on the determination of 223 rotation measures from polarization observations of relatively distant southern pulsars made using the Parkes radio telescope. Combined with previously published observations these data give clear evidence for large-scale counterclockwise fields (viewed from the north Galactic pole) in the spiral arms interior to the Sun and weaker evidence for a counterclockwise field in the Perseus arm. However, in interarm regions, including the Solar neighbourhood, we present evidence that suggests that large-scale fields are clockwise. We propose that the large-scale Galactic magnetic field has a bisymmetric structure with reversals on the boundaries of the spiral arms. Streaming motions associated with spiral density waves can directly generate such a structure from an initial inwardly directed radial field. Large-scale fields increase toward the Galactic Center, with a mean value of about 2~$\mu$G in the Solar neighbourhood and 4~$\mu$G at a Galactocentric radius of 3 kpc., Comment: ApJ accepted. 16 pages, 14 figures, 2 tables, 223 pulsar RMs

A new drifting pulsar, PSR J0815+09, was discovered in the Arecibo drift-scan searches. An intriguing feature of this source is that within the four pulse components in the integrated pulse profile, the sub-pulse drifting direction in the two leading components is opposite from that in the two trailing components. In view that the leading theoretical model for pulsar sub-pulse drifting can only interpret one-direction sub-pulse drifting, the observed "bi-drifting" phenomenon from PSR J0815+09 poses a great challenge to the pulsar theory. The inner annular gap (IAG), a new type of inner particle accelerator, was recently proposed to explain both $\gamma$-ray and radio emission from pulsars. Here we show that the coexistence of the IAG and the conventional inner core gap (ICG) offers a natural interpretation to the bi-drifting phenomenon. In particular, the peculiar drifting behavior in PSR J0815+09 can be reproduced within the inverse Compton scattering (ICS) model for pulsar radio emission., Comment: 11 pages, 2 figures, ApJL, accepted