In the context of complex and dynamic marine environment, the offloading of computing tasks for ships of Internet of Things (IoT) users is a very challenging problem considering the different quality of service (QoS) requirements of maritime applications. Mobile edge computing driven by powerful computing capability and edge intelligence is taken as a promising solution, especially for the resource-constrained and delay-sensitive maritime IoT users. In this paper, we study the optimal edge server selection problem for ship IoT users to jointly minimize the latency and energy consumption for task offloading. Specifically, we first propose a novel space-air-ground-edge (SAGE) integrated maritime network architecture to offload computation-intensive IoT services at sea. Then, the latency and energy consumption of data transmission and processing during offloading are modelled. Based on the models, the edge server selection problem is formulated into a Multi-Armed Bandits learning problem, with considering the task latency requirement and energy budget. To achieve the optimal solution, a novel algorithm, referred to as UCB1-ESSS, is developed, which links the latency, energy consumption, and network constraints by introducing both reward and cost. The simulation results show that the proposed algorithm can achieve considerably lower offloading latency and weighted latency-energy cost compared with the traditional algorithms under different QoS requirements, which proves the efficacy of theproposed algorithm.

The scheduling of data access plays a prominent part in a memory-based fast Fourier transform (FFT) processor. In this article, a conflict-free data-access scheme is proposed and applied to implement a radix-2^k FFT processor with the computing parallelism being the power of 2. The proposed FFT module exploits single-port RAMs to retain the area efficiency, and it has the following hardware-friendly features. First, the memories are merged into four banks under the arbitrary computing parallelism, which helps to circumvent the complicated data-routing circuit. Second, the parallelism of the input/output (I/O) data can be enlarged to the internal computing parallelism, which accelerates the data update and enhances the throughput capability. Furthermore, the concurrent I/O operation that overlaps the storage of the FFT input with the supply of FFT output in the same memory sets is allowed, which minimizes the memory occupation for the management of continuous data flow. Based on the testing on the field-programmable gate array (FPGA) and ASIC implementation using the SMIC 40-nm CMOS technology, it is demonstrated that the proposed FFT processor can offer high throughput and simultaneously achieve remarkable area and power efficiency.

Rapid and accurate rendezvous and proximity operations for spacecraft are crucial to the success of most space missions. In this paper, a sequential convex programming method, combined with the first-order and second-order Birkhoff pseudospectral methods, is proposed for the autonomous rendezvous and proximity operations of spacecraft. The original nonlinear and nonconvex close-range rendezvous problem with thrust constraints and no-fly zone constraints is converted into its convex version by using the sequential convexification techniques; then, the Birkhoff pseudospectral method is used to transcribe the dynamic constraints into a series of linear algebraic equality constraints, in other words, a convex second-order conic programming problem with a relatively small condition number. Thus, the resulting problem can be accurately and efficiently solved by a convex solver. The simulation results indicate that the proposed methods, especially the second-order Birkhoff pseudospectral method, have obvious advantages over other methods in computational efficiency and sensitivity.