2024 | Ivan Zryianoff, Lorenzo Gigli, Federico Montori, Luca Sciullo, Carlos Kamien ski, Marco Di Felice
CACHE-IT is a distributed framework for proactive edge caching in heterogeneous IoT scenarios, designed to address the unique requirements of IoT environments. It decouples the caching strategy from the underlying architecture, enabling customization based on application-specific needs. The framework allows for flexible deployment, modification, and replacement of caching strategies, and provides clients' request history as input to the caching strategy. CACHE-IT components are based on current Web technology standards, ensuring compatibility and easy integration with existing systems.
The framework is designed to handle the dynamic nature of IoT environments, including the volatile nature of sensors and the need for interoperability between heterogeneous devices. It utilizes cooperative and proactive caching mechanisms to optimize resource sharing and reduce latency. CACHE-IT also supports a twofold validation approach, including extensive simulations and real-world deployment in a Structure Health Monitoring (SHM) system.
The contributions of CACHE-IT include flexibility in caching strategies, IoT-oriented design with a dedicated device abstraction layer, an advanced caching mechanism that supports cooperative caching, and twofold validation through simulations and real-world deployment. The framework is evaluated through large-scale simulations and real-world deployment in a real IoT environment, demonstrating its effectiveness in reducing latency, improving hit rates, and minimizing the number of requests sent to data providers.
CACHE-IT is implemented using industry-adopted applications and custom components, with the Cache Manager invoking caching strategy functions through a POST request. The framework is designed to be deployed using lightweight virtualization, such as Docker containers. The cache storage is implemented using Redis, a lightweight key-value store that operates entirely in memory, making it suitable for low-latency use cases. The framework is validated through simulations and real-world deployment, demonstrating its flexibility, applicability, and performance in IoT scenarios.CACHE-IT is a distributed framework for proactive edge caching in heterogeneous IoT scenarios, designed to address the unique requirements of IoT environments. It decouples the caching strategy from the underlying architecture, enabling customization based on application-specific needs. The framework allows for flexible deployment, modification, and replacement of caching strategies, and provides clients' request history as input to the caching strategy. CACHE-IT components are based on current Web technology standards, ensuring compatibility and easy integration with existing systems.
The framework is designed to handle the dynamic nature of IoT environments, including the volatile nature of sensors and the need for interoperability between heterogeneous devices. It utilizes cooperative and proactive caching mechanisms to optimize resource sharing and reduce latency. CACHE-IT also supports a twofold validation approach, including extensive simulations and real-world deployment in a Structure Health Monitoring (SHM) system.
The contributions of CACHE-IT include flexibility in caching strategies, IoT-oriented design with a dedicated device abstraction layer, an advanced caching mechanism that supports cooperative caching, and twofold validation through simulations and real-world deployment. The framework is evaluated through large-scale simulations and real-world deployment in a real IoT environment, demonstrating its effectiveness in reducing latency, improving hit rates, and minimizing the number of requests sent to data providers.
CACHE-IT is implemented using industry-adopted applications and custom components, with the Cache Manager invoking caching strategy functions through a POST request. The framework is designed to be deployed using lightweight virtualization, such as Docker containers. The cache storage is implemented using Redis, a lightweight key-value store that operates entirely in memory, making it suitable for low-latency use cases. The framework is validated through simulations and real-world deployment, demonstrating its flexibility, applicability, and performance in IoT scenarios.