Human Activity Recognition (HAR) underpins applications in healthcare, rehabilitation, fitness tracking, and smart environments, yet existing deep learning approaches demand dataset-specific training, large labeled corpora, and significant computational resources. We introduce RAG-HAR, a training-free retrieval-augmented framework that leverages Large Language Models (LLMs) for HAR. RAG-HAR computes lightweight statistical descriptors, retrieves semantically similar samples from a vector database, and uses this contextual evidence to make LLM-based activity identification. We further enhance RAG-HAR by first applying prompt optimization and introducing an LLM-based activity descriptor that generates context-enriched vector databases for delivering accurate and highly relevant contextual information. Along with these mechanisms, RAG-HAR achieves state-of-the-art performance across six diverse HAR benchmarks. Most importantly, RAG-HAR attains these improvements without requiring model training or fine-tuning, emphasizing its robustness and practical applicability. RAG-HAR moves beyond known behaviors, enabling the recognition and meaningful labelling of multiple unseen human activities.

Recent cyberattacks have increasingly targeted distributed networking environments like IoT networks. To detect these attacks, hidden under network traffic encryption, many centralized Machine Learning (ML) based solutions have been introduced, which are not well suited for IoT networks. This work proposes PIFL a practical approach to secure IoT networks by combining federated learning, in-network ML using P4-enabled devices, software-defined networks, and binarized neural networks. PIFL detects compromised edge devices and isolates them into separate network slices based on trust parameters derived from their behavior. We demonstrate the feasibility of PIFL using an experimental testbed with three intelligent network devices and seven IoT devices implemented on Raspberry Pi devices.

Immense growth of network usage and the associated proliferation of network, traffic, traffic classes, and diverse QoS requirements pose numerous challenges for network operators. Though data-driven approaches can provide better solutions for these challenges, limited data has been a barrier to developing those methods with high resiliency. In this work, we propose SyNIG (Synthetic Network Traffic Generation through Time Series Imaging), which utilizes Generative Adversarial Networks (GANs) for network traffic synthesis by converting time series data to a specific image format called GASF (Gramian Angular Summation Field). With GASF images we encode correlation between samples in 1D signals on a single 2D pixel map. Taking three types of network traffic; video streaming, accessing websites and IoT, we synthesize over 200,000 traces using over 40,000 original traces generalizing our method for different network traffic. We validate our method by demonstrating the fidelity of the synthetic data and applying them to several network related use cases showing improved performance.

360° videos are getting popular providing an immersive streaming experience for the user, nevertheless, demand high bandwidth in mobile networks due to their larger spherical frames. In this preliminary work, we propose to combine frame interpolation and super resolution methods to optimize tile based 360° video delivery by streaming them at low qualities in network and increasing the quality leveraging Multi Access Edge Computing. We propose a mechanism to adaptively decide this quality conversion at the client side which improves average video quality by 30% and bandwidth saving by 43.3% compared to existing tile based streaming.

360° videos have become popular creating an immersive streaming experience for the user, nevertheless, these videos demand high bandwidth in operational networks and have strict latency requirements. Viewport (VP) aware streaming with variable tiling has been proposed as a promising solution to reduce bandwidth consumption while providing fine granularity to the user VP. Content caching at the edge has also been proposed to reduce the delivery latency. Though, combining these two mechanisms have potential advantages, applying conventional tile-based caching, which primarily tries to find identical tiles, is not feasible due to the high diversity in tile area and location in variable tiles. To this end, we propose OpCASH, an ILP based mechanism to devise optimal cache tile configuration at a MEC server to provide a non-overlapping tile cover for a given VP request in variable tiles, while minimizing the requests to remote servers and reducing the delivery latency. Experimental trace-driven simulation results show that we can achieve more than 95% of VP coverage from cache after just 24 views of the video. Compared to a baseline which represents conventional tile-based caching, OpCASH reduces the data fetched from the content servers by 85% and total content delivery time by 74%.

360° videos a.k.a. spherical videos are getting popular among users nevertheless, omnidirectional view of these videos demands high bandwidth and processing power at the end devices. Recently proposed viewport aware streaming mechanisms can reduce the amount of data transmitted by streaming a limited portion of the frame covering the current user viewport (VP). However, they still suffer from sending a high amount of redundant data, as the fixed tile mechanisms can not provide a finer granularity to the user VP. Though, making the tiles smaller can provide a finer granularity for user viewport, it will significantly increase encoding-decoding overhead. To overcome this trade-off, in this paper, we present a computational geometric approach based adaptive tiling mechanism named VASTile, which takes visual attention information on a 360° video frame as the input and provides a suitable non-overlapping variable size tile cover on the frame. Experimental results show that VASTile can save up to 31.1% of pixel redundancy before compression and 35.4% of bandwidth saving compared to recently proposed fixed tile configurations, providing tile schemes within 0.98 (±0.11) seconds time frame.

Unlike conventional videos, 360° videos give freedom to users to turn their heads, watch and interact with the content owing to its immersive spherical environment. Although these movements are arbitrary, similarities can be observed between viewport patterns of different users and different videos. Identifying such patterns can assist both content and network providers to enhance the 360° video streaming process, eventually increasing the end-user Quality of Experience (QoE). But a study on how viewport patterns display similarities across different video content, and their potential applications has not yet been done. In this paper, we present a comprehensive analysis of a dataset of 88 360° videos and propose a novel video categorization algorithm that is based on similarities of viewports. First, we propose a novel viewport clustering algorithm that outperforms the existing algorithms in terms of clustering viewports with similar positioning and speed. Next, we develop a novel and unique dynamic video segment categorization algorithm that shows notable improvement in similarity for viewport distributions within the clusters when compared to that of existing static video categorizations.

Streaming 360° video demands high bandwidth and low latency, and poses significant challenges to Internet Service Providers (ISPs) and Mobile Network Operators (MNOs). The identification of 360° video traffic can therefore benefits fixed and mobile carriers to optimize their network and provide better Quality of Experience (QoE) to the user. However, end-to-end encryption of network traffic has obstructed identifying those 360° videos from regular videos. As a solution this paper presents 360NorVic, a near-realtime and offline Machine Learning (ML) classification engine to distinguish 360° videos from regular videos when streamed from mobile devices. We collect packet and flow level data for over 800 video traces from YouTube & Facebook accounting for 200 unique videos under varying streaming conditions. Our results show that for near-realtime and offline classification at packet level, average accuracy exceeds 95%, and that for flow level, 360NorVic achieves more than 92% average accuracy. Finally, we pilot our solution in the commercial network of a large MNO showing the feasibility and effectiveness of 360NorVic in production settings.

360°/VR videos are getting popular. However, these videos demand high bandwidth and processing power at the end devices. Though, viewport (VP) aware streaming can reduce the amount of data transmitted by streaming a limited portion of the frame covering the current user viewport, popular content providers still transfer the entire panoramic frame which demands more bandwidth. Also, these mechanisms, which partition the frames into a fixed number of tiles can not provide a finer boundary to cover the user VP, causing high pixel redundancy. To address these issues, first, we propose an offline and near-realtime 360° vs normal video classification tool which is further extended to analyse 360° video streaming distribution in the wild. Secondly, we propose a content aware 360° video partitioning tool leveraging a computational geometric approach. Our initial results show the feasibility of both proposals.

Unlike the traditional Internet application such as web browsing and peer-to-peer(P2P), video streaming has been dominating the global network traffic for the past few years, raising many challenges for network providers. With the popularity of interactive videos, a.k.a 360° videos, resource requirement for video streaming has been further increased. Prior identification of these video traffic is useful for effective provisioning of network resources, yet it is difficult due to the end-to-end encryption of data. However, with the recent advances in Machine Learning (ML) methods, prior identification of these resource-demanding traffic types has become viable. Nonetheless, they require more training data, without which leads to poor performance. Collecting more training data may also pose issues related to delayed training time. To remedy this problem, in this paper, we propose a novel Generative Adversarial Network (GAN) based data generation solution to synthesise video streaming data targeting 360°/normal video classification. Taking over 600 actual video traces and generating ≈ 30000 new traces, our post-classification results show that we can achieve 5 - 15% of accuracy improvement compared to only having actual traces.

The security and privacy of the end-users are a few of the most important components of a communication network. Though end-to-end encryption (e.g., TLS/SSL) fulfils this requirement, it makes inspecting network traffic with legacy solutions such as Deep Packet Inspection difficult. Recent Machine Learning techniques have shown outstanding performance in encrypted traffic classification. Nevertheless, such approaches require efficient flow sampling at real enterprise-scale networks due to the sheer volume of transferred data. Through this paper, we propose a holistic architecture to extract flow information of encrypted data at multi Gbps line rate using sampling and sketching mechanisms, enabling network operators to estimate flow size distribution accurately and understand the behavior of VPN-obfuscated traffic. Using over 6000 video traffic traces, under three main evaluation scenarios based on trace duration and starting time point, we show that it is possible to achieve 99% accuracy for service provider classification and over 90% accuracy for content classification for a given service provider in the best case. We also deploy our solution at an operational enterprise-scale network leveraging kernel bypassing to demonstrate its capability to efficiently sample live traffic for analytics.

While end-to-end encryption brings security and privacy to the end-users, it makes legacy solutions such as Deep Packet Inspection ineffective. Despite the recent work in machine learning-based encrypted traffic classification, these new techniques would require, if they were to be deployed in real enterprise-scale networks, an enhanced flow sampling due to sheer volume of data being traversed. In this paper, we propose a holistic architecture that can cope with encryption and multi-Gbps line rate with sampling and sketching flow statistics, which allows network operators to both accurately estimate the flow size distribution and identify the nature of VPN-obfuscated traffic. With over 6000 video traffic traces, we show that it is possible to achieve 99% accuracy for service provider classification even with sampled possibly inaccurate data.