In active direction of arrivals (DOA) sensing, the multiple-input multiple-output (MIMO) radar can lead to more degrees of freedom (DOFs). The DOFs can be further increased when the fourth order cumulants are used in MIMO. However, most array geometries are not designed for fourth order cumulants based MIMO and this array geometry optimization problem is still an open issue. To have a simple guideline for this optimization problem, we reformulate it into two separate problems, which are a fourth order sum coarray optimization problem and a second order difference coarray problem. The fourth order sum coarray optimization is realized by solving a postage stamp problem and a virtual nested array can be obtained. The number of elements in this virtual nested array can be significantly increased with only a few physical sensors. The corresponding inter-element spacing of virtual nested array is expanded simultaneously. Then the second order difference coarray is adopted to the virtual nested array to further increase the DOFs. Numerical results validate the high DOFs achieved by our proposition.
25 February 2025 in IEEE Transactions on Instrumentation and Measurement
In this paper, a modified ESPRIT algorithm is firstly introduced for the estimation of direction-of-arrival (DOA), eliminating the necessity for a prior knowledge of the number of signal sources. Eigenvalue decomposition (EVD) is no longer necessary for the construction of the signal subspace, substantially curtailing computational demands, and the procedure is modified specifically according to the characteristics of the new signal subspace. Secondly, to further improve the computational efficiency, the real-valued implementation is considered based on an improved unitary transformation, which can maintain the estimation accuracy. Simulation results affirm that the refined ESPRIT-like algorithm matches the computational precision of the conventional ESPRIT methodology, even in the absence of preliminary source enumeration. Notably, the execution of this improved algorithm on a Field-Programmable Gate Array (FPGA) reveals its capability to expedite DOA estimation within an abbreviated latency, all the while maintaining a thrifty usage of computational resources. This demonstration underscores the efficiency and feasibility of the proposed method for real-time applications demanding rapid and resource-efficient DOA computations without the source number.
Lidar sensors play a pivotal role in a multitude of remote sensing domains, finding extensive applications in various sectors, including robotics, unmanned aerial vehicles (UAVs), autonomous driving, and 3D reconstruction, among others. Their significance and versatility in these areas have made them indispensable tools for a wide range of applications. The accuracy of Lidar odometry (LO), which serves as the front end of SLAM, is crucial. In this paper, we propose a novel iterative closest point (ICP) technique that combines semantic features to improve LO precision. First, the semantic segmentation neural network is used to extract the semantic features from each frame of the point cloud. Then, the obtained semantic features assist in extracting the local geometric features of the point cloud. Also, the residual blocks of the ICP algorithm’s least squares are combined with semantic confidence functions to better predict the exact pose. Compared to LOAM, there is an average improvement of 4 cm in accuracy per 100 m. Experimental results illustrate the superiority of the proposed method and indicate that the fusion of semantic features can robustly improve the precision of LO.