In-band network telemetry (INT) enables real-time network monitoring by embedding telemetry data into packets. The advent of programmable switches further enhances the flexibility of INT by enabling dynamic customization of telemetry collection at the hardware level. However, the practical application of INT is hampered by significant challenges arising from data noise (due to packet loss, delay, and measurement inaccuracies) and network dynamics (such as changing INT paths and feature requirements). These issues severely degrade the performance of machine learning models used for analyzing INT data, hindering the accurate capture of spatio-temporal network characteristics. This paper presents Planner, a novel generative graph learning framework designed to address these limitations. Planner enables the collection of network features at various levels of granularity on programmable switches. Crucially, it constructs dynamic graphs representing evolving INT paths and employs a hybrid Graph Neural Network (GNN) and Recurrent Neural Network (RNN) architecture to effectively learn spatial and temporal dependencies. Furthermore, Planner incorporates variational inference to generate robust latent representations, mitigating the detrimental effects of noise and instability in INT data. We have implemented a testbed prototype of Planner using Intel Tofino ASIC switches. Extensive experiments demonstrate the performance superiority and robustness of Planner over the baseline methods, achieving a 23.2% improvement in F1 score.

The aim of this study was to investigate the effect of different loading symbols and loading durations on online video viewers’ perception of waiting time. Sixty young adults participated in this study and their subjective ratings on waiting time perception for forty-eight different loading symbols (4 progress functions × 2 shapes × 2 embellishments × 3 durations) were obtained through a 7-point Likert scale. Each participant chose the best loading symbols in terms of waiting time perception at the end of the experiment. Results showed that duration and the progress function practically affect the viewers’ waiting time perception, and shape and embellishment do not. Loading symbols of power and inverse power progress functions are perceived as shorter than those of the repetitive and linear progress functions. It is recommended to show loading progress and use manipulated progress functions, and design factors such as shape and embellishment are considered to be less important for designing a better loading symbol. The findings of this study may serve as a useful input for video service providers and symbol designers in creating shortly perceived loading symbols.