The statistical properties of the repeating fast radio burst FRB 121102 are investigated. We find that the cumulative distributions of fluence, flux density, total energy and waiting time can be well fitted by the bent power law. In addition, the probability density functions of fluctuations of fluence, flux density and total energy well follow the Tsallis $q$-Gaussian distribution. The $q$ values keep steady around $q\sim 2$ for different scale intervals, indicating a scale-invariant structure of the bursts. The statistical properties of FRB 121102 are very similar to that of the soft gamma repeater SGR J1550-5418. The underlying physical implications need to be further investigated.

Axion monodromy is a generic phenomenon in string compactifications. We investigate the production of gravitational-wave background from axion monodromy inflation during preheating. By performing lattice simulations using pseudo-spectral methods, we find that significant single-peak stochastic gravitational-wave backgrounds are generated during preheating for asymptotic linear $|\phi|$ and $\phi^{2/3}$ potentials in axion monodromy inflation and cuspy potentials. This may allow us to explore string compactifications through gravitational waves.

We constrain three cosmological models, i.e. the concordance cold dark matter plus a cosmological constant ($\Lambda$CDM) model, Power-law (PL) model, and $R_h=ct$ model using the available local probes, which includes the JLA compilation of type-Ia supernovae (SNe Ia), the direct measurement of Hubble constant (H(z)), and the baryon acoustic oscillations (BAO). For $\Lambda$CDM model, we consider two different cases, i.e. zero and non-zero spatial curvature. We find that by using the JLA alone, it is indistinguishable between $\Lambda$CDM and PL models, but the $R_h=ct$ model is strongly disfavored. If we combine JLA+H(z), the $\Lambda$CDM model is strongly favored against the other two models. The combination of all the three datasets also supports $\Lambda$CDM as the best model. We also use the low-redshift ($z<0.2$) data to constrain the deceleration parameter using cosmography method, and find that only the $\Lambda$CDM model is consistent with cosmography. However, there is no strong evidence to distinguish between flat and non-flat $\Lambda$CDM models by using the local data alone.

We propose a framework of Finsler space-time to explain the observed parity asymmetry and the power deficit in the low-$\ell$ ($2\leqslant \ell \leqslant 29$) multipole range of cosmic microwave background (CMB) temperature anisotropies. In the $3+1$ dimensional space-time, the three-dimensional space is described by a Randers-Finsler space, which is spatially irreversible, inducing the parity asymmetry in the CMB angular power spectrum. We estimate the constraints on the two parameters introduced by Finsler space-time via analyzing the low-$\ell$ angular power spectrum in PLANCK 2015 CMB temperature data. We see that the low-$\ell$ power suppression in the CMB temperature anisotropies can also be resolved in this scenario. Our study shows that the two low-$\ell$ anomalies, i.e., parity asymmetry and power deficit, may have a common origin.

We test the cosmological principle by fitting a dipolar modulation of distance modulus and searching for an evolution of this modulation with respect to cosmological redshift. Based on a redshift tomographic method, we divide the Joint Light-curve Analysis compilation of supernovae of type Ia into different redshift bins, and employ a Markov-Chain Monte-Carlo method to infer the anisotropic amplitude and direction in each redshift bin. However, we do not find any significant deviations from the cosmological principle, and the anisotropic amplitude is stringently constrained to be less than a few thousandths at $95\%$ confidence level.

In this work, we study a direction dependent power spectrum in anisotropic Finsler space-time. We use this direction dependent power spectrum to address the low-l power observed in WMAP and PLANCK data. The angular power spectrum of the temperature fluctuations has a lower amplitude in comparison to the $\Lambda$CDM model in the multipole range l = 2-40. Our theoretical model gives a correction to the isotropic angular power spectrum $C^{TT}_l$ due to the breaking of the rotational invariance of the primordial power spectrum. We estimate best-fit model parameters along with the six $\Lambda$CDM cosmological parameters using PLANCK likelihood code in CosmoMC software. We see that this modified angular power spectrum fits the CMB temperature data in the multipole range l=2-10 to a good extent but fails for the whole multipole range l = 2-40.

The statistical properties of the soft gamma repeater SGR J1550--5418 are investigated carefully. We find that the cumulative distributions of fluence, peak flux and duration can be well fitted by a bent power law, while the cumulative distribution of waiting time follows a simple power law. In particular, the probability density functions of fluctuations of fluence, peak flux, and duration have a sharp peak and fat tails, which can be well fitted by a $q$-Gaussian function. The $q$ values keep approximately steady for different scale intervals, indicating a scale-invariant structure of soft gamma repeaters. Those results support that the origin of soft gamma repeaters is crustquakes of neutron stars with extremely strong magnetic fields.

Einstein's equivalence principle (EEP) can be tested by the time delay between photons with different energies passing through a gravitational field. As one of the most energetic explosions in the Universe, gamma-ray bursts (GRBs) provide an effective tool to test the accuracy of EEP. In this paper, we use the continuous spectra of 20 short GRBs detected by the Swift/BAT to test the validity of EEP. Taking the duration of GRBs as the upper limit of the time delay induced by EEP violation (assuming that the high energy photons arrive later than the low energy photons), the difference of the parameterized post-Newtonian parameter is constrained with high accuracy. The strictest constraint, $|\gamma(150~{\rm keV})-\gamma(15~{\rm keV})|<5.59\times 10^{-10}$ from GRB 150101B, is about $1\sim 2$ orders of magnitude tighter than previous constraints. Moreover, our result is more statistically significant than previous results because we use the continuous spectra instead of isolated photons.

In quantum gravity, a foamy structure of space-time leads to Lorentz invariance violation (LIV). As the most energetic astrophysical processes in the Universe, gamma-ray bursts (GRBs) provide an effective way to probe quantum gravity effects. We use continuous spectra of 20 short GRBs detected by the Swift satellite to give a conservative lower limit of quantum gravity energy scale $M_\textrm{QG} $. Due to the LIV effect, photons with different energy have different velocities. This will lead to the delayed arrival of high energy photons relative to the low energy ones. Based on the fact that the LIV-induced time delay can't be longer than the duration of a GRB, we present the most conservative estimation of the quantum gravity energy scales from 20 short GRBs. The most strict constraint, $M_\textrm{QG}>5.05\times10^{14}$ GeV, is from GRB 140622A.