Abstract The multi‐mode heterogeneous network combines the advantages of high‐speed power line communication (HPLC) and high radio frequency (HRF), ensuring service quality and meeting the requirements for data transmission delay and reliability even when devices are flexibly deployed. Queuing delay and privacy entropy are important metrics for managing multi‐mode heterogeneous internet of things (IoT) networks, which require collaborative optimization of the transmission phase (server selection and sub‐channel allocation) and the computing phase (computing resource allocation) to ensure low latency and high privacy entropy. However, existing communication‐computing collaborative optimization methods face issues such as low privacy security of electricity‐carbon service data, high difficulty in solving the joint optimization problem, and resource competition. Therefore, this paper proposes a privacy preservation‐driven communication‐computing collaboration method for the management of multi‐mode heterogeneous IoT networks. Firstly, the architecture for the management of multi‐mode heterogeneous IoT networks is constructed and a privacy entropy model for electricity‐carbon computing service data is established to measure the privacy security performance of the network management. Secondly, a joint optimization problem of queuing delay and privacy entropy under long‐term privacy entropy constraints are constructed and the long‐term privacy entropy constraints from short‐term decisions is decoupled based on Lyapunov optimization. Finally, a joint optimization algorithm for server selection and multi‐mode sub‐channel allocation driven by privacy protection is proposed. This algorithm reduces the three‐dimensional matching optimization problem among different devices, servers, and channels, and uses auction matching to solve the conflict of resource block selection, further optimizing the computing resource allocation of edge servers based on the Karush–Kuhn–Tucker (KKT) conditions. Simulation results show that the proposed algorithm effectively reduces queuing delay and improves privacy security of data transmission.

The understanding and representation of songs is a crucial issue in music platforms, as it can facilitate numerous applications in the music field. Songs are a common multimodal art form within the music domain, achieving rich musical connotations and strong expressiveness. However, the data of songs exhibit obvious multimodal and heterogeneous characteristics, presenting significant challenges to the understanding and representation of songs. Regrettably, the current methods do not respond to these challenges effectively. To this end, in this study, a unified multimodal representation framework for songs, namely SUMMR, is proposed. Specifically, first, a two-layer framework is put forward. In embedding layer, the features of different modalities data are embedded into a unified space. In content layer, a novel cross-modal attention mechanism is designed, which effectively capture the cross-modal semantic associations and deep music features, thereby obtaining a unified representation of songs. Then, a two-level hierarchical pre-training algorithm is proposed, which can effectively lower the training cost. Finally, experiments are conducted on two typical music tasks with public datasets of songs, where the experimental results demonstrate the effectiveness of SUMMR for understanding and representation of songs, and also show that SUMMR has good capability of being fine-tuned in many song-based tasks.

A novel intermittent dynamic event-triggered control is proposed to investigate the exponential synchronization in mean square (ESMS) of stochastic complex networks (SCNs). In contrast to existing literature, the proposed intermittent control is based on dynamic event-triggering instead of static time-triggering during the control interval. Meanwhile, the number of event triggers can be reduced by introducing an exponential function. Moreover, for each sample path solution of SCNs, the existence of a minimum positive inter-event time is guaranteed in all the event-generators proposed in this paper. In addition, based on the Lyapunov method and the graph theory, synchronization criteria for the ESMS of SCNs under the intermittent dynamic event-triggered control are established. Finally, theoretical results are applied to islanded microgrid systems and simulations are given to verify the effectiveness of the results.

In this paper, we present and implement a novel procedure for the accurate simulation of the thermoelectric effects which determine micromachined structures' behavior. This approach is simple to discretize, with rapid numerical convergence properties, and it yields results which are in very good agreement with experimental observations. We employ this procedure in particular to calculate P/sub 1ATM//P/sub VAC/, where P/sub 1ATM/ (resp. A/sub VAC/) denotes heat conducted down the micromachined structure supporting arms at one atmosphere (resp. vacuum) when the device is at the same average temperature. We show that P/sub 1ATM//A/sub VAC/ may not be unity, as had been earlier assumed, but is independent of input current. Finally, for a special class of devices, an analytical formula is obtained for P/sub 1ATM//P/sub VAC/ in terms of device parameters, and this result is also compared with the numerical simulation results.

We have developed a software package that simulates the operation of a silicon micromachined CMOS thermal conductivity gas pressure gauge. The performance of actual devices was compared against the simulated operation and was found to be in good agreement. The 3-dimensional simulation was reduced to two 2-dimensional simulations to reduce complexity. The two equations resulting from steady state energy balance considerations were discretized and an iterative nonlinear Gauss-Seidel procedure applied to solve the system of equations. Temperature profiles and contours were calculated and the effect of geometric and materials modifications was demonstrated.>