Modern artificial intelligence (AI) workloads are characterized by increasing complexity and large datasets with multi-billions of parameters, which place unique demands on the computational and communication infrastructure. These requirements expose the limit of server-centric AI systems in scalability and responsiveness, particularly for tasks operating on high-volume network data flows. While serverless architectures offer promising alternatives through on-demand execution and adaptive scaling, they often struggle with compute-intensive and latency-sensitive AI workloads. To address these challenges, we propose a novel adaptive in-network execution framework for distributed AI inference tasks that adopts information-centric networking (ICN) principles to enable data-centric orchestration and seamless service discovery for latency-sensitive intelligent services. Our proposal treats AI tasks as named invocable functions, which allow computational requests (e.g., object detection, anomaly analysis) to be seamlessly routed, resolved, and executed on intermediate network nodes based on intent-aware resolution and in-network availability. The experimental results show that the proposed system achieves a minimum task completion latency and maintains a success rate of over 96% across all tasks.

The continuous growth of the Internet of Things (IoT) and its expanding network drives an exponential increase in data generation, which necessitates robust and efficient processing solutions for latency-sensitive applications. While local processing on resource-constrained devices can be inefficient, offloading computations to third parties introduces significant security and privacy risks. Homomorphic encryption (HE) offers a powerful solution for secure computations on encrypted data without compromising confidentiality. However, the computational intensity of HE results in high latency, limiting its applicability in real-time decision-making scenarios. To overcome these challenges, we propose ShieldDINC, a distributed in-network computation scheme designed for efficiency and privacy protection using localized edge processing to optimize secure implementations with HE. In addition, it leverages the efficient data delivery and caching capabilities inspired by an information-centric networking (ICN) approach. ShieldDINC ensures that sensitive data is securely offloaded and computed at the nearest edge router, which caches results for potential reuse, reducing retrieval time and repetitive computations. Extensive simulations demonstrate that ShieldDINC achieves a 30% latency reduction and a 20% computational efficiency improvement for real-time intensive computations compared to state-of-the-art alternatives while promising robust security and privacy.

Information-centric networking (ICN) brings in-network caching as a significant characteristic along with its novel network architecture. In-network caching serves as an efficient content distribution strategy in which intermediate routers store and redistribute data to reduce network latency, congestion, and retrieval time. Meanwhile, it decouples contents from the publisher, and content integrity and confidentiality turn out to be problematic. In this study, we propose a mechanism named “in-network secure content management” (ISCM) to facilitate content integrity and authenticity in ICN by exploiting identity-based cryptography (IBC) for content signature verification. In addition, it guarantees content confidentiality by applying a hybrid encryption-based access control mechanism to prevent unauthorized content access and tampering attacks. Evaluation results reveal that ISCM offers secure content retrieval in ICN with a scalable key distribution without significant computational overhead and communication delay.

In today’s healthcare landscape, emerging technologies serve as crucial foundations, and the integration of blockchain technology into this rapidly evolving digital framework is invaluable. However, the healthcare industry has long struggled with managing the immense and ever-expanding amount of big data electronic health records (EHRs) collected from various sources, including the Internet of Things (IoT), wearable devices, and mobile applications. The storage of all these big data EHRs on the chain creates blockchain bloat, leading to slow transaction speed and high storage costs. More importantly, it incurs security and privacy leakage problems while sharing patients’ sensitive data transparently with a wide range of users in a distributed network. To address these challenges, we propose a secure and privacy-preserving data management system for distributed off-chain networks called PrivOff based on blockchain technology, which stores big EHRs separately in the decentralized file system with efficient access control enforcement and flexible revocation to prevent data breaches. In addition, patients can securely share data without revealing their unique identities to ensure privacy. The evaluation results and security analysis reveal that PrivOff can notably reduce the blockchain storage burden while offering data security, patient privacy, and high data availability.

The newly emerging technologies and applications primarily focus on content distribution over the Internet, and a current host-centric TCP/IP Internet architecture becomes infeasible to fulfill this demand. Named Data Networking (NDN) is one of the most promising Future Internet architecture that facilitates a content-centric communication model over TCP/IP architecture. NDN supports in-network caching as the main characteristic that provides efficient scalability and minimum latency of content retrieval. Each NDN router possesses a content store table to cache the contents and directly serve the same requests in the future. Content can be cached anywhere in the NDN network, and content security and confidentiality become vital to prevent content access by unauthorized consumers. A hybrid encryption-based access control scheme has been proposed to address content confidentiality concerns by applying symmetric and identity-based proxy re-encryption schemes in NDN. However, it still requires further implementation and evaluation analyses with related schemes to prove that it offers a lower computational overhead and faster content retrieval time while protecting content confidentiality in NDN architecture. This paper conducts an additional experimental study on the scheme and shows some evidence about the reduction of the computational and communication time.