By decoupling compute and memory resources into independent pools that are provisioned and managed separately, disaggregated memory (DM) is promising to break the scaling constraints for memory systems and improve resource utilization. However, it also comes with a new challenge to design a high-performance hashing index to manage the vast memory pool with weak computing power. In this paper, we reconsider this problem and find that existing hashing indexes for DM still experience two fundamental yet unresolved limitations: (i) amplifying traffic under high insertion concurrency, and (ii) introducing significantly high insertion latency, stemming mainly from the directory synchronization and item relocation in resizing process. We resolve the above limitations by designing Pulse, a finegrained hierarchical hashing index for DM. Pulse comprises the following three design primitives. It proposes a multi-level index structure, which breaks the conventional flat directory into multiple sub-directories that are organized hierarchically, achieving fine-grained directory synchronization. Pulse also maintains a small portion of hashed keys in the memory pool, which aids in item relocation during resizing, thereby reducing resizing traffic and promising system stability. Pulse finally exploits operation parallelism by tailoring the doorbell batching mechanism with selective signaling. We conduct extensive experiments using a variety of benchmarks, showing that Pulse can improve$3.46 \times$of the throughput and reduce 76.3% of the tail latency compared to state-of-the-art hashing indexes.

ABSTRACT Motivated by real‐world applications in logistics and post‐disaster relief, this study introduces a novel variant of the orienteering problem, termed the orienteering problem with heterogeneous drones (OP‐HD). In this problem, a single truck collaborates with multiple heterogeneous drones to serve a subset of customers. While the truck is stationed at a customer location, multiple drones can be launched simultaneously, each capable of performing multiple trips. The objective is to maximize the total collected profit within a limited time horizon. We formulate the OP‐HD as a mixed‐integer linear programming model and discuss potential extensions to accommodate broader application scenarios. To solve the problem efficiently, we develop an adaptive large neighborhood search (ALNS) heuristic with customized destroy and repair operators. Computational experiments based on classical Solomon benchmark instances are conducted to evaluate the effectiveness of the proposed ALNS algorithm, assess the advantages of the profit‐oriented OP‐HD model, and analyze the impact of key parameter settings. The results demonstrate that the truck‐drone cooperative system can serve more customers and achieve higher profits within a limited time horizon compared to a truck‐only system. Furthermore, the advantage of employing heterogeneous drones becomes more significant in scenarios where customers are spatially clustered. The results further show that time is the primary operational constraint, and that improving battery capacity, energy efficiency, and drone speed enhances system profitability. Finally, a case study based on delivery areas in North Carolina is conducted to demonstrate the practical applicability of the proposed model.

Persistent memory (PMem) brings new design considerations in realizing high-performance and scalable hashing indexes. We uncover that existing hashing indexes for PMem still suffer from traffic amplification and memory inefficiency. We present MetoHash, a memory-efficient and traffic-optimized hashing index on hybrid PMem-DRAM memories. MetoHash proposes a three-layer index structure spanning across CPU caches, DRAM, and PMem for data management. It aggregates the incoming key-value items in CPU caches for fast inserts, which are then arranged in DRAM and flushed to PMem, to eliminate traffic amplification. MetoHash also uses fingerprinting to reduce unnecessary probings over PMem and removes duplicate items during bucket relocations. We implement MetoHash on PMem with persistent and volatile CPU caches, and show that compared to state-of-the-art hashing indexes for PMem, MetoHash improves the throughput by 86.1–257.6% under various workloads.

This article presents a linear motor mover position detection method based on an image sample database to achieve high-precision measurement of the position of a long-stroke linear motor mover. First, the aperiodic fringe image is constructed with a chirp signal as the target shooting source, and an image sample database is established by capturing the target image with a certain acquisition frequency using a line-scan camera. Second, a normalized correlation coefficient (NCC) matching method is proposed to search for the position of the current image in the database. A coarse positioning method based on velocity estimation is also employed to improve the efficiency of position matching in the sample database. Third, to overcome the position measurement error caused by incomplete matching between the current image and the image in the database, a subpixel measurement algorithm based on local upsampling NCC is proposed to improve the measurement accuracy. Finally, the actual displacement of the actuator is obtained by combining the calibration coefficients. To verify the feasibility of the proposed method, an experimental platform is built using a line-scan camera, a linear motor, an adjustable light source, and a target source image. Comparison tests, speed adaptation experiments, and light adaptation experiments are set up. The experimental results show that the average time of the proposed method is approximately 1 ms, and the average absolute measurement error is approximately 0.005 mm under various working conditions, which shows that the proposed method has real-time performance and strong environmental adaptability.

To improve the tracking accuracy and antidisturbance performance of permanent magnet linear synchronous motor (PMLSM), in this article, a fast recursive terminal sliding-mode control (FRTSMC) with mover velocity and disturbance observer (MVDO) is proposed. The FRTSMC is designed by a recursive structure of fast terminal sliding mode and integral sliding mode. The FRTSMC is initialized with an appropriate value to reduce the sliding-mode approach time and guarantee that the tracking error converges to zero in a finite time. To further improve the antidisturbance property and suppress the chattering phenomenon, an MVDO is integrated into the FRTSMC to simultaneously estimate the mover velocity and the lumped disturbance. Lyapunov theory is employed to analyze the closed-loop stability of the composited controller. The experimental results verify the effectiveness of the proposed method in reducing convergence error and enhancing robustness compared with the conventional fast nonsingular TSMC, adaptive recursive TSMC, and nonlinear disturbance observer based FRTSMC for PMLSM.

By providing non-volatility with DRAM-comparable performance, the emerging persistent memory (PMem) is propelling new key-value (KV) store designs. The recently released Intel Optane DC PMem now shifts the persistent boundary from memory up to CPU caches, which further eliminates the needs of cacheline flush instructions used in extensive KV stores. However, we uncover via testbed experiments that this change can even degrade the performance of existing KV stores once directly deploying them atop the new generation of the Optane PMem, stemming mainly from the mismatch of access granularities and heavy software designs.In this paper, we present CacheKV, the first KV store built atop persistent CPU caches. CacheKV allocates per-core sub-MemTable in CPU caches with a lazy index update mechanism, so as to fast absorb incoming writes. It then proposes a copy-based flush mechanism to convert small-sized cacheline evictions into large-sized flushes to suppress the write amplification. CacheKV finally accelerates read operations via periodically compacting the sub-skiplists. Extensive testbed experiments show that CacheKV improves the write throughput by 19.5× on average in the write-dominated environment without compromising the read performance, when compared to the state-of-the-art KV stores for the PMem.

To improve the bandwidth and thrust performance of permanent magnet synchronous linear motor (PMSLM), from the perspective of motor structure design, this article proposes a high-accuracy and fast-calculation inductance model of centralized winding based on analytic kernel-embedded elastic-net regression (AKER) to establish a comprehensive analytical model of PMSLM performance. A stochastic fractal search (SFS) algorithm is introduced to optimize the comprehensive performances of PMSLM. First, Neumann's formula of integral is used to establish the equivalent inductance analytical model (EIAM) for the three-phase winding of PMSLM, and the analytic kernel is constructed based on the EIAM. In combination with the finite element simulation samples of inductance, the analytical kernel is embedded into the elastic-net regression to establish the inductance model of high efficiency, namely AKER, and compared with the finite element mode (FEM), EIAM and other advanced methods, the superiority of AKER is verified from the accuracy and time cost. Then, based on the proposed AKER inductance model, the current and voltage equations of PMSLM are decoupled in the two-phase rotating coordinate system to establish the comprehensive analysis model of the PMSLM performance. Based on the comprehensive model, the optimization function of PMSLM is constructed, and the SFS algorithm is introduced to solve the function iteratively. Finally, the effectiveness of the proposed modeling and optimization method is verified by FEM-based control simulation and prototype experiment, respectively.