The frequency-dependent periodic active window of the fast radio burst FRB 180916.J0158+65 (FRB 180916B) was observed recently. In this Letter, we propose that a Be/X-ray binary (BeXRB) system, which is composed of a neutron star (NS) and a Be star w
ith a circumstellar disk, might be the source of a repeating FRB with periodic activities, and apply this model to explain the activity window of FRB 180916B. The interaction between the NS magnetosphere and the accreted material results in evolution of the spin period and the centrifugal force of the NS, leading to the change of the stress in the NS crust. When the stress of the crust reaches the critical value, a starquake occurs and further produces FRBs. The interval between starquakes is estimated to be a few days that is smaller than the active window of FRB 180916B. When the NS moves out of the disk of the Be star, the interval between starquakes becomes much longer than the orbital period, which corresponds to the non-active phase. In this model, due to the absorption of the disk of the Be star, a frequency-dependent active window would appear for the FRBs, which is consistent with the observed properties of FRB 180916B. And the contribution of dispersion measure (DM) from the disk of the Be star is small. In addition, the location of FRB 180916B in the host galaxy is consistent with a BeXRB system.
The first repeating fast radio burst (FRB), FRB 121102, was found to be associated with a spatially coincident, persistent nonthermal radio source, but the origin of the persistent emission remains unknown. In this paper, we propose that the persiste
nt emission is produced via synchrotron-heating process by multiple bursts of FRB 121102 in a self-absorbed synchrotron nebula. As a population of bursts of the repeating FRB absorbed by the synchrotron nebula, the energy distribution of electrons in the nebula will change significantly. As a result, the spectrum of the nebula will show a hump steadily. For the persistent emission of FRB 121102, the total energy of bursts injecting into the nebula is required to be about $3.3times10^{49},unit{erg}$, the burst injection age is over $6.7times 10^4,unit{yr}$, the nebula size is $sim0.02,unit{pc}$, and the electron number is about $3.2times10^{55}$. We predict that as more bursts inject, the brightness of the nebula would be brighter than the current observation, and meanwhile, the peak frequency would become higher. Due to the synchrotron absorption of the nebula, some low-frequency bursts would be absorbed, which may explain why most bursts were detected above $sim1~unit{GHz}$.
