Digital twin (DT) is envisioned not only to perform the high-fidelity virtual representation of its corresponding physical entity (PE), but also to serve as an active agent delivering diverse types of sophisticated services. This paper studies an end-edge-cloud collaborative DT placement and update framework. Specifically, we consider that DTs are dynamically placed across edge servers (ESs) via migration following their paired PEs' potential mobility, while being supported by real-time data fetched from the cloud center and user ends. On top of this, we emphasize a unique feature that DTs should also be continually updated capturing the uncertain evolutions for both personalized service ability improvement and versatile service ability maintenance, where the personalization is improved by utilizing the experiential knowledge from the cloud center and their corresponding PEs, and the versatility is maintained by integrating pre-stored profiles. To maximize the long-term system-wide average weighted quality-of-service (QoS) in handling all types of PEs' service requests under the stringent system cost constraint, we formulate an online problem to jointly optimize DT migrations, service priorities towards various request types, and all related DT updating strategies. To address underlying difficulties, we propose a novel adversarial bandit learning assisted online optimization approach, called ARBOK. We first leverage the Lyapunov decomposition method to transform the long-term problem into multiple instant ones, each of which is further decoupled into two correlated subproblems. For solving one subproblem with a bilinear structure, we develop a McCormick envelopes based algorithm (MO-EL). Besides, we design an extended adversarial combinatorial multi-armed bandit algorithm (AC-BL) to tackle the other subproblem, which constructs a super arm set to resolve the issue of excessively large decision space and employs a robust scheme to handle the inherent uncertainty and non-stationarity in each super arm's loss function. We integrate both algorithms seamlessly into ARBOK and alternately execute them till the convergence. Theoretical analysis and extensive simulations show the effectiveness of the introduced dynamic DT placement and continual update framework, demonstrating that ARBOK can converge to the asymptotic optimum within a polynomial-time complexity while outperforming counterparts.

Socially Assistive Robotics (SAR) has shown promise in supporting emotion regulation for neurodivergent children. Recently, there has been increasing interest in leveraging advanced technologies to assist parents in co-regulating emotions with their children. However, limited research has explored the integration of large language models (LLMs) with SAR to facilitate emotion co-regulation between parents and children with neurodevelopmental disorders. To address this gap, we developed an LLM-powered social robot by deploying a speech communication module on the MiRo-E robotic platform. This supervised autonomous system integrates LLM prompts and robotic behaviors to deliver tailored interventions for both parents and neurodivergent children. Pilot tests were conducted with two parent-child dyads, followed by a qualitative analysis. The findings reveal MiRo-E’s positive impacts on interaction dynamics and its potential to facilitate emotion regulation, along with identified design and technical challenges. Based on these insights, we provide design implications to advance the future development of LLM-powered SAR for mental health applications.

This paper presents a combinatorial online learning framework for the reliable deployment of containerized Digital Twin (DT) systems in mobile edge computing, addressing challenges such as unpredictable edge server failures and variable writable-layer sharing permissions. By jointly optimizing read-only layer placement, writable layer creation, and cross-server layer loading while adhering to long-term latency and energy constraints, the framework enhances DT service reliability. We first decouple the original problem into a series of deterministic subproblems via Lyapunov optimization, and then propose a contextual bandit mechanism to explore the unknown layer sharing permission information. Theoretical analysis establishes regret bounds, while simulation experiments validate the effectiveness of the proposed framework.

In this paper, we study a mobility-aware two-timescale online optimization for constructing a digital twin (DT) -assisted task execution system under end-edge-cloud collaboration. DTs are deployed on edge servers deployed on roadside units (RSUs), and needs to be proactively migrated based on the future location of its corresponding vehicle. To minimize the task response latency executed by DT with stringent energy consumption constraint, we jointly optimize the uploading frequency of vehicle’s status data for adaptive mobility prediction, the DT migration decision along with the communication and computation resource allocations. Considering that decision variables are triggered asynchronously, we propose a novel two-timescale mobility-aware online optimization approach (TMO), which first employs an extended two-timescale Lyapunov method to decompose the problem into a series of instant subproblems and then integrates a multi-armed bandit (MAB) algorithm to dynamically determine uploading frequency of vehicle’s status data as the length of the large-timescale. After that, for each small-timescale problem, we develop a GRU-based vehicle trajectory preidiction method to predict the location of vehicles, followed by an alternate minimization (AM) based algorithm to decide remaining decision variables based on the predicted trajectory. Theoretical analyses and simulations show that the proposed approach can reach asymptotic optimum and demonstrate its superiority over counterparts.

This paper presents a low-power multimode eight-channel analog front-end (AFE) circuit for direct time-of-flight (dToF) LiDAR applications. The proposed AFE features rich programmability in: a) Two operation modes, parallel mode and selectable mode to save power; b) Two coupling modes, anode and cathode coupling with photodiodes (PDs); c) Two gain modes, high-gain mode for PIN-PD and low-gain mode for APD. Correspondingly, several circuit design techniques have been proposed to support the programmability including: a) Reconfigurable TIA with awake/sleep state and high/low gain mode switching; b) Symmetrical pulse width control to match the pulse width of anode and cathode coupling and input over load protection; 3) Glitch reduction for channel switchover in selectable mode. Designed in 0.18-µm CMOS technology, the AFE achieves bandwidth of 180 MHz, transimpedance gain of 100 dBW, input-referred noise current of 2.1 ${\text{pA}}/\sqrt {{\text{Hz}}} $ Hz and output swing of 1 Vpp,diff. The channel switchover time is around 10 ns in selectable mode. The channel power consumption is 50 mW in parallel mode and 11.4 mW in selectable mode.

Container–based microservice provisioning, with its elasticity in terms of the layered structure, enables the sharing of common layers among different edge computing tasks, both within and across edge servers (ES). However, due to the potential hardware breakdowns, each ES may prone to failures, affecting its lifetime (i.e., the time-length that an ES works continuously without interruptions), and in turn leading to the collapse of their hosted/provided microservices or the other ESs’ microservices requesting common layers from it. To address such an issue, in this paper, we study the microservice deployment optimization with layer sharing for maximizing the system-wide reliability while satisfying all tasks’ delay requirements. Considering dynamic task generations and the asynchronization of various decision variables with different triggers, we design an online optimization algorithm by leveraging an improved Lyapunov technique integrating randomized rounding, Lagrangian method and convex optimization, which iteratively solves the problem over different timescales. Theoretical analyses and simulations evaluate the performance of the proposed solution, showing that it can achieve an increase of 12.4% in reliability and a reduction of 28.57% in total delay, compared to the counterparts.

Due to the potential of revolutionizing a variety of human-centric services, human digital twin (HDT) is envisioned to become an important part of our daily life. The HDT applications need to frequently collect and process data obtained from individuals and their environment, analyzing each physical twin while updating its corresponding virtual twin, which will consume a large amount of computing, storage and sensing resources cumulatively. Meanwhile, running HDT applications, such as emotion recognition, naturally contains the executions of several dependent tasks. Considering the resource limitations of mobile terminals, we enable dependent task offloading to mitigate terminal load and reduce the latency of HDT applications. Specifically, this paper proposes an energy and cost-aware offloading algorithm for dependent tasks with edge-cloud collaboration to empower HDT applications. We show that the problem of dependent task offloading under constraints of service cost and terminal energy consumption is NP-hard. The complexity of task interdependency makes the offloading decision under dual constraints even more challenging. The proposed offloading algorithm firstly generates task paths based on task interdependency and computation load, deriving the initial solution. Then, task reassignment and CPU frequency scaling methods are utilized to further optimize the obtained solution. Simulation results illustrate that our approach can achieve better performance in terms of makespan and service success ratio compared to the existing approaches.

Human digital twin (HDT) is an emerging paradigm that bridges physical twins (PTs) with powerful virtual twins (VTs) for assisting complex task executions in human-centric services. In this paper, we study a two-timescale online optimization for building HDT under an end-edge-cloud collaborative framework. As a unique feature of HDT, we consider that PTs' corresponding VTs are deployed on edge servers, consisting of not only generic models placed by downloading experiential knowledge from the cloud but also customized models updated by collecting personalized data from end devices. To maximize task execution accuracy with stringent energy and delay constraints, and by taking into account HDT's inherent mobility and status variation uncertainties, we jointly and dynamically optimize VTs' construction and PTs' task offloading, along with communication and computation resource allocations. Observing that decision variables are asynchronous with different triggers, we propose a novel two-timescale accuracy-aware online optimization approach (TACO). Specifically, TACO utilizes an improved Lyapunov method to decompose the problem into multiple instant ones, and then leverages piecewise McCormick envelopes and block coordinate descent based algorithms, addressing two timescales alternately. Theoretical analyses and simulations show that the proposed approach can reach asymptotic optimum within a polynomial-time complexity, and demonstrate its superiority over counterparts.