For detection and estimation of 2-D frequencies from a 2-D array of data using a subspace decomposition method, one needs to construct a block Hankel matrix. For reliable detection andestimation, the rank of the block Hankel matrix should be made equal to the number of 2-D frequenciesinherent in the data in the absence of noise. In this work, we provide the conditions for achieving the desired rank

The control of the spin-Seebeck current is still a challenging task for the development of spin caloritronic devices. Here, we construct a spin-Seebeck device by inserting a strongly correlated quantum dot (QD) between the metal lead and magnetic insulator. Using the slave-particle approach and non-crossing approximation, we find that the spin-Seebeck effect increases significantly when the energy level of the QD locates near the Fermi level of the metal lead due to the enhancement of spin flipping and occurrences of quantum resonance. Since this can be easily realized by applying a gate voltage in experiments, the spin-Seebeck device proposed here can also work as a thermovoltaic transistor. Moreover, the optimal correlation strength and the energy level position of the QD are discussed to maximize the spin-Seebeck current as required for applications in controllable spin caloritronic devices.

Subspace computation is fundamental for many signal processing applications. A well-known tool for computing the principal subspace of a data matrix is the power method. During the iterations of the power method, a proper normalization is essential to avoid numerical overflow or underflow. Normalization is also needed to achieve desirable properties such as orthonormalized subspacematrices. A number of normalization techniques for the power method is reviewed, which include the conventional as well as nonconventional ones. In particular, a new method of normalization is introduced to achieve asymptotical orthonormalization of subspace matrices without the use of squareroot. This method is among a class of normalization methods that allow a simple adaptive implementation of the power method for subspace tracking.

Jet-medium interactions are studied via a multi-hadron correlation technique (called "2+1"), where a pair of back-to-back hadron triggers with large transverse momentum is used as a proxy for a di-jet. This work extends the previous analysis for nearly-symmetric trigger pairs with the highest momentum threshold of trigger hadron of 5 GeV/$c$ with the new calorimeter-based triggers with energy thresholds of up to 10 GeV and above. The distributions of associated hadrons are studied in terms of correlation shapes and per-trigger yields on each trigger side. In contrast with di-hadron correlation results with single triggers, the associated hadron distributions for back-to-back triggers from central Au+Au data at $\sqrt{s_{NN}}$=200 GeV show no strong modifications compared to d+Au data at the same energy. An imbalance in the total transverse momentum between hadrons attributed to the near-side and away-side of jet-like peaks is observed. The relative imbalance in the Au+Au measurement with respect to d+Au reference is found to increase with the asymmetry of the trigger pair, consistent with expectation from medium-induced energy loss effects. In addition, this relative total transverse momentum imbalance is found to decrease for softer associated hadrons. Such evolution indicates the energy missing at higher associated momenta is converted into softer hadrons.