Abstract Machine unlearning in the context of cybersecurity and privacy protection facilitates the removal of specific training data impacts from deep learning (DL) models, adhering to security, privacy, or compliance demands. However, traditional methods can only handle short-term, independent unlearning tasks. Conversely, real-world scenarios often involve extensive unlearning demands from users. Current methods fail to adequately address these demands due to substantial computational overhead and adverse impacts on inference accuracy, leaving the security and privacy of many users at risk. To navigate these challenges adeptly, we introduce the Multi-Agent Reinforcement Learning Data Lifecycle Management (MADLM) strategy. MADLM intricately examines the interactions between unlearning and continuous learning processes, enabling the postponement of certain tasks for combined execution to optimize computational resources. Concurrently, it employs strategic data management to maintain and enhance inference accuracy. Furthermore, by utilizing Multi-Agent Reinforcement Learning (MARL), MADLM dynamically orchestrates task scheduling to minimize computational demands, improve task response times, and bolster inference reliability, crucial for upholding stringent cybersecurity and privacy standards. Our evaluations of MADLM reveal substantial enhancements, including a 6% uplift in inference accuracy and a dramatic reduction in computational overhead to merely 12% of the original demands, effectively expanding the data security protections.

Within the Internet of Things (IoT) landscape, Encoder as a Service (EaaS) is a cloud-based service for many AI empowered scenarios, such as autopilot and face-scan payment, enabling IoT devices to keep a light classifier model at local while remotely accessing powerful encoding models. However, adversarial examples like an image with imperceptible tiny perturbations that can lead to incorrect classifying results, make it challenging to achieve a robust EaaS-based classifier. Certified radius (R) emerges as a measure against such imperceptible tiny perturbations, and guarantees the trustworthiness of any input with perturbations smaller than R. Despite offering R related services, the substantial computational overhead from traditional methods, stemming from the extensive searches of R for each request, poses a barrier to their practical deployment. In this article, we identify the repetitive computations in the EaaS framework due to the fixed encoding model and propose a novel search cache scheme to speed up the R-related computation. Therefore, we propose large-scale efficient robust EaaS (ScaleRES) to address different R-related services: First, ScaleRES strategically stores previous computation results of R, to enable the reuse and refining of R across users. Then, ScaleRES obtains the average certified radius (ACR) efficiently with selective cached R. Finally, ScaleRES performs R filtering to enhance adversarial training for a robust EaaS-based classifier. Comprehensive evaluations demonstrate that in three distinct computations, ScaleRES offers significant savings in computational overhead compared to the conventional approach—70% for R computation as the number of clients increases, 35% for ACR of the entire test set, and 40% for adversarial training with robustness comparable to other works.

Securing the communication of wireless devices within the Internet of Things (IoT) ecosystem is a critical concern that garners considerable attention. Nowadays, leveraging physical layer data to generate lightweight channel keys for wireless devices has become a prevalent approach. However, most existing methodologies focus primarily on the key establishment for singular communication instances, often neglecting the intricate needs for ongoing key generation and renewal throughout multiple communication sessions. This paper introduces a novel channel key regeneration scheme that leverages previously utilized information from successful channel key generations. To enhance the randomness and security of the renewed key, the proposed design incorporates a linear feedback shift register (LFSR) for the generation and adjustment of new channel keys. This method aims to establish a more dynamic and secure channel key management framework, catering to the evolving demands of IoT communications.

Physical layer channel key generation is a promising technology for secure communication in wireless network security , which mainly establishes secure communication keys between any two legitimate users. This article reviews current techniques for physical layer channel key generation. The physical layer channel key generation principle, system model, attack model, key generation process, and experimental application are comprehensively reviewed. In addition, we introduce the experimental scenarios of key generation and the collection methods of channel measurements in wireless applications and simulation platforms and summarize the experimental equipment configuration and actual operation in the process of collecting different measurements. The article concludes with some suggestions for future research.

Device-to-device communication is increasingly important for emerging wireless applications, including the Internet of Things and industry mobile networks. To establish a secure communication channel between multiple legitimate devices, key agreement is an essential first step. Physical layer channel information is a promising technology for achieving communication security by generating a shared channel key. In this paper, we propose an efficient and random method for generating channel keys based on channel state information (CSI). Firstly, we preprocess the available subcarriers to reduce the correlation of adjacent subcarriers. Secondly, we propose a novel quantization process to generate random and secure channel keys. The part consists of two phases: the quantization phase and the inconsistency removal phase. In the quantization stage, CSI sub carriers are grouped into small blocks, and appropriate thresholds are generated. Then, wireless users quantize the CSI subcarriers into initial bit streams of 0 and 1 with different thresholds. In the inconsistency removal phase, we design an error correction mechanism where wireless users exchange the discarded index sequences in the initial bit streams to achieve key sequence consistency. We implement the proposed scheme using off-the-shelf wireless devices and evaluate its performance. The experimental results show that each CSI packet can produce 30 bits, which have passed the NIST randomness tests.

Recently, deep learning (DL) becomes increasingly important since its transformative effect on a wide range of applications. During inference process, the DL model is deployed on the cloud to answer online queries. One crucial issue in the progress of DL inference is energy consumption, which significantly retards computation performance. Therefore, many previous investigations decrease the energy consumption via computation reuse technique based on similarity. However, if input data consists individually from mobile devices, applying these schemes will significantly decline computation performance. Because in disordered individual inputs, similarity for reuse is difficult to exploit directly. Results of initial experimental observations show that (1) individual input data also has high similarity for reuse, and (2) the total similarity during computation process has a relation with the characteristics of input data. This motivates us to design a reordering scheme to enhance similarity for computation reuse. Our main approaches are using statistical theory to predict the similarities among input data, and determining the execution sequence. Based on these approaches, we propose an effective input reordering scheme for computation reuse to save energy consumption. The evaluation under various benchmarks demonstrates that the reordering scheme significantly outperforms the previous schemes, for instance, the computation reuse is enhanced to $1.1 \times$ and the energy consumption is minimized to 40% according to the configuration of traditional computation reuse technique.

Parameter Server (PS) has been widely used to train a large amount of data on multiple machines in parallel. In parameter server, a critical problem is how to effectively schedule multiple training jobs to minimize the job completion time. Some existing work has proposed methods of setting the number of concurrent workers. However, they do not effectively consider the topology of GPU placement which affects the efficiency of communication. This paper proposes a novel resource-to-time model based on the number of workers and the topology of GPU placement. According to the model, we propose an algorithm called TOPO-PS particularly for topology problem in parameter servers. The algorithm achieves the placement strategy based on graph mapping algorithm. Evaluation under various algorithms evidences the superiority of our algorithm. TOPO-PS yields shorter job completion, by up to 53.48% of that of FIFO and 88.77% of OASIS.