Integrated sensing and communication (ISAC) is a promising technique for establishing an effective Internet of Things (IoT) network, utilizing the same hardware and resource block for both sensing and communication functions. In this study, we propose a hybrid demodulation scheme-enabled ISAC beamforming design to enhance sensing capabilities. Our study delves into the combined effects of diverse users’ demodulation schemes and sensing power constraints on the design of transmit beamforming for multi-user communication and multi-target sensing. We formulate three optimization problems for beamforming design which tackle user-target discrimination, user-target overlap, and a general hybrid scenario. In particular, in the hybrid scenario, we incorporate the type of communication receivers as an optimized variable and introduce additional sensing power constraints to enhance sensing performance in terms of beampattern and time waveform without compromising communication performance. To effectively tackle these optimization challenges, we introduce a hierarchical convex relaxation framework considering user-target pairing. Simulation results illustrate that the proposed beamforming approach improves sensing performance while maintaining communication performance, surpassing traditional designs under different user-target directional distributions.

Reconfigurable intelligent surfaces (RIS) have emerged as a crucial technology for making sustainable and green communication in future sixth-generation (6G) networks. By adaptively adjusting the wireless environment, RIS facilitates flexible control over signal propagation. When RIS is integrated into a simultaneous wireless information and power transfer (SWIPT) system, it can improve energy efficiency (EE) and link reliability. This enables low-power internet-of-things (IoT) devices to achieve continuous energy acquisition while maintaining reliable data access. This capability aligns well with the increasing demands of future 6G networks for enhancing EE and achieving ubiquitous connectivity. Motivated by this, we investigate a RIS-assisted multi-user SWIPT system that accounts for the nonlinear characteristics of the high power amplifier (HPA) at the transmitting end and the nonlinearity of the energy harvesting (EH) circuits at the receivers. These hardware imperfections pose major challenges for system modeling and optimization, particularly under stringent power limitations and quality of service (QoS) requirements, where improving overall EE is a key objective. Therefore, we proposed a joint optimization framework that simultaneously designs the access point (AP) beamforming, the RIS reflection coefficients, and the power splitting (PS) ratios at the receivers. The resulting problem is non-convex, with strong coupling among these variables. To tackle this issue, we propose an efficient alternating optimization (AO) framework. The fractional EE objective is handled using the Dinkelbach method, while each subproblem is iteratively convexified via successive convex approximation (SCA) and semi-definite relaxation (SDR) algorithms. Simulation results reveal that the proposed AO approach achieves substantial EE gains and confirm that integrating RIS conspicuous boosts the overall system EE performance.

The log-structured nature of NAND flash storage necessitates garbage collection in SSDs. Garbage collection (GC) is a major source of runtime write amplification (WA), leading to faster device wear out and interference with host I/Os. The key to mitigating this problem is separating data by lifetime so that data in the same flash block are invalidated within temporal proximity. For higher lifetime prediction accuracy and adaptibility, prior works proposed using machine learning algorithms for data separation. However, existing learning-based solutions perform data lifetime prediction at the host side, leading to several drawbacks. First, host-side prediction does not have knowledge of the internal data movement inside the SSD during GC, and thus fails to leverage the opportunity to further separate GC writes, resulting in suboptimal WA reduction in the long term. Second, performing prediction at the host significantly prolongs the I/O critical path and consumes host resources that could otherwise be used for serving user applications. We present Shiro, a holistic FTL design that performs instorage data separation for both user writes and GC writes for maximal long-term WA reduction. For user writes, Shiro uses a sequence model to accurately predict data lifetime by learning lifetime distribution from long historical access patterns. For GC writes, Shiro incorporates a reinforcement learning-assisted page migration strategy that takes direct feedback from longterm WA to further improve data separation efficacy. To address the challenges posed by performing fine-grained and real-time machine learning decisions inside the resource-constrained SSD, we propose a suite of enabling techniques to keep computation and storage overhead low. Extensive evaluation of Shiro on real-world traces shows that Shiro can deliver 29 WA compared with conventional FTL and state-of-the-art instorage data separation schemes. Furthermore, thanks to lower data migration overhead during GC, Shiro achieves significantly higher steady-state I/O performance.

Regular path queries (RPQs) in graph databases are bottlenecked by the memory wall. Emerging processing-in-memory (PIM) technologies offer a promising solution to dispatch and execute path matching tasks in parallel within PIM modules. We present an efficient PIM-based data management system tailored for RPQs and graph updates. Our solution, called PimBeam, facilitates efficient batch RPQs and graph updates by implementing a PIM-friendly dynamic graph partitioning algorithm. This algorithm effectively addresses graph skewness issues while maintaining graph locality with low overhead for handling RPQs. PimBeam streamlines label filtering queries by adding a filtering module on the PIM side and leveraging the parallelism of PIM. For the graph updates, PimBeam enhances processing efficiency by amortizing the host CPU's update overhead to PIM modules. Evaluation results of PimBeam indicate 3.59x speedup for RPQs and 29.33x speedup for graph update on average over the state-of-the-art traditional graph database.

Regular path queries (RPQs) in graph databases are bottlenecked by the memory wall. Emerging processing-in-memory (PIM) technologies offer a promising solution to dispatch and execute path matching tasks in parallel within PIM modules. We present Moctopus, a PIM-based data management system for graph databases that supports efficient batch RPQs and graph updates. Moctopus employs a PIM-friendly dynamic graph partitioning algorithm, which tackles graph skewness and preserves graph locality with low overhead for RPQ processing. Moctopus enables efficient graph update by amortizing the host CPU's update overhead to PIM modules. Evaluation of Moctopus demonstrates superiority over the state-of-the-art traditional graph database.

In the Internet of Everything (IoE) scenarios, the extensive deployment of devices may result in more stringent power and communication needs. Within this context, we utilize the reconfigurable intelligent surface (RIS) to support the simultaneous wireless information and power transfer (SWIPT) system, whereby the stable transmission of energy and information services can be guaranteed. Specifically, we construct the system model through electromagnetics (EMs), which is based on the scattering-parameter (S-parameter) analysis, for revealing the crucial factors of the practical hardware. Relying on the model, the energy-efficient (EE) maximization problem constrained to the Quality of Services (QoS) is proposed for the users with the framework of co-located receiver (Rx). However, the problem is more intractable due to the introduced channel model. To resolve it, we propose an effective optimization scheme. First, the Neuman series approximation method is adopted to deconstruct the EM transfer model. Then the reformed problem, which includes the variables (i.e., the power splitting ratio, the active beamformer, and the reflection-coefficient matrix), can be addressed through the strategy of alternative optimization (AO). Further, the inner convex approximation (INCA) scheme and Dinkelbach’s algorithm are applied to tackle each subproblem. In the numerical simulation, we demonstrate that the array configuration can influence not only the hardware properties of RIS but also the EE performance of the whole system. What is more, the proposed scheme performs better for the tightly coupled RIS owing to the awareness of the mutual-coupling (MC) effect.

Pointer chasing becomes the performance bottleneck for today's in-memory indexes due to the memory wall. Emerging processing-in-memory (PIM) technologies are promising to mitigate this bottleneck, by enabling low-latency memory access and aggregated memory bandwidth scaling with the number of PIM modules. Prior PIM-based indexes adopt a fixed granularity to partition the key space and maintain static heights of skiplist nodes among PIM modules to accelerate index operations on skiplist, neglecting the changes in skewness and hotness of data access patterns during runtime. In this article, we present RADAR, an innovative PIM-friendly skiplist that dynamically partitions the key space among PIM modules to adapt to varying skewness. An offline learning-based model is employed to catch hotness changes to adjust the heights of skiplist nodes. In multiple datasets, RADAR achieves up to 198.2x performance improvement and consumes 47.4% less memory than state-of-the-art designs on real PIM hardware.

Garbage collection in SSDs causes write amplification. The key to mitigating this problem is separating data by lifetime. Prior works proposed using machine learning to accurately predict data lifetime but prediction is performed at the host side, burdening the host storage stack. We present PHFTL, a practical, holistic FTL design with device-side learning-based data separation. The machine learning model in PHFTL accurately and adaptively predicts the lifetime of every written page. A suite of enabling techniques are introduced to keep computation and storage overhead low. Extensive evaluation of PHFTL demonstrates superiority over state-of-the-art and feasibility on real hardware.

In this paper, we propose a signal design unconstrained ISAC framework, where different transmit signal design has no significant effect on both sensing and communication performance. To begin with, we describe the model of transmit signals with transmission power limitations. Next, the sensing and communication performance are analysed with reasonable metrics respectively. The property of unconstrained signal design is implied in a proved signal intensity distribution theorem, which facilitates the sensing function of our framework. Then, a beamforming optimization problem is formulated to maximize a weighted transmission rate subject to the guaranteed sensing performance. The formulated non-convex problem is approximately solved by alternatively solving two simpler sub-problems. Finally, numerical results are given to validate the effectiveness, practicality and satisfactory performance of our proposed framework.

The simultaneous wireless information and power transfer (SWIPT) technology assisted by the reconfigurable intelligent surface (RIS) can bring flexibility and stability to the end nodes of the internet of things (IoT) during the deployment. In this paper, we propose a RIS-aided SWIPT system based on a hardware transfer model from the electromagnetic perspective. Particularly, an energy efficiency (EE) maximization problem subject to the quality of service (QoS) demands, power resource budget and circuit restrains is introduced. Furthermore, the active beamforming vectors of the BS and the circuit parameters at the RIS are optimized jointly. The problem can be decomposed into two sub-problems and solved iteratively until convergence. In particular, semi-definite relaxation (SDR), successive convex approximation (SCA), Dinkelbach's algorithm are applied to the solutions of the sub-problems. Numerical results reveal the influences of the various QoS requirements on EE performance. Moreover, the actual generated beams of the BS and the RIS are shown to demonstrate the effectiveness of the proposed optimization strategy.

The widespread deployment of the Internet of Things (IoT) is promoting interest in simultaneous wireless information and power transfer (SWIPT), the performance of which can be further improved by employing a reconfigurable intelligent surface (RIS). In this article, we propose a novel RIS-enhanced SWIPT system built on an electromagnetic-compliant framework. The mutual-coupling effects in the whole system are presented explicitly. Moreover, the reconfigurability of RIS is no longer expressed by the reflection-coefficient matrix but by the impedances of the tunable circuit. For comparison, both the no-coupling and the coupling-awareness cases are discussed. In particular, the energy efficiency (EE) is maximized by cooperatively optimizing the impedance parameters of the RIS elements as well as the active beamforming vectors at the base station (BS). For the coupling-awareness case, the considered problem is split into several subproblems and solved alternatively due to its nonconvexity. First, it is transformed into a more solvable form by applying the Neuman series approximation, which can be resolved iteratively. Then, an alternative optimization (AO) framework and semidefinite relaxation (SDR), successive convex approximation (SCA), and Dinkelbach’s algorithm are applied to solve each subproblem decomposed from it. Owning to the similarity between the two cases, the no-coupling one can be viewed as a reduced form of the coupling case and, thus, solved through a similar approach. Numerical results reveal the influence of mutual-coupling effects on the EE, especially in the RIS with closely spaced elements. In addition, physical beam designs are presented to demonstrate how the RIS assists SWIPT through various reflecting states in different conditions.