Hierarchical Cloud Computing Architecture for Context-Aware IoT Services

Hierarchical Cloud Computing Architecture for Context-Aware IoT Services
Abstract
This paper presents a new cloud computing model for context-aware Internet of Things (IoT) services. The proposed computing model is hierarchically composed of two layers: a cloud control layer (CCL) and a user control layer (UCL). The CCL manages cloud resource allocation, service scheduling, service profile, and service adaptation policy from a system performance point of view. Meanwhile, the UCL manages end-to-end service connection and service context from a user performance point of view. The proposed model can support nonuniform service binding and its real-time adaptation using meta-objects. Furthermore, it supports intelligent service-context management using a supervised and reinforcement learning-based machine learning framework. We implemented a lightweight prototype of the proposed computing model. Evaluations confirm that the proposed computing model offers enhanced performance compared with legacy uniform computing models.
EXISTING SYSTEM :
Recently, many context-aware cloud computing platforms have been proposed. presented a new geo-distributed cloud computing infrastructure for ad-hoc mobile network users, called MobiCloud. It supports internet-based resource outsourcing by allowing context-aware mobile users to construct a virtual network system through virtualization and programmable networking technologies. It can provide ad-hoc mobile users with reduced service access latency and increased cloud infrastructure utilization. presented content-centric cloud computing architectures to improve the content delivery performance. In particular, proposed content delivery framework with Software Defined Networking (SDN) and Content Centric Networking (CCN). Additionally to serve autonomic optimal services, it proposed a reinforcement learning-based context-aware content delivery scheme. proposed a service replication scheme together with a selfconguration approach for the activation and hibernation of the replicas of a service depending on relevant context information from the mobile system. To that end, an election algorithm has been designed and implemented. presented a mobility-efficient cloud computing architecture. It proposed multi-homed mobile IP (M-MIP) as a multi-homed mobility management protocol to facilitate soft, low latency handoffs with minimal packet loss. M-MIP enables a user to connect to several access networks simultaneously before initiating the handoff process.
The user periodically probes the registered network interfaces to select a target network for handoff without disconnecting from the previous network interface, thereby minimizing network delay and packet losses during the handoff process. presented off-loading efficient cloud computing architectures. They proposed context-aware off-loading controls obtaining the lowest task execution time and energy consumption for the off-loadable tasks. The proposed control decides when it is benecial to off-load, which wireless medium is used for off-loading and which resources to use as the off-loading location considering a set of context parameters, multiple wireless medium and mobile cloud resources. Presented verifiable and secure cloud computing architecture to enhance the trust of cloud computing. It first proposed full homomorphic encryption technologies to process data in an encrypted form at Cloud Service Provider (CSP) in order to protect the privacy of data providers and data owners. It further deployed an auditing protocol to verify the correctness of encrypted data processing by applying a Trusted Auditing Proxy .
PROPOSED SYSTEM :
• The proposed platform provides a non-uniform servicebinding model that provides application-specific computing environments at the platform-level as shown in Fig. 2(b), rather than at the user level.
• The proposed platform supports real-time adaptable service binding that consists of application and transmission bindings. It supports binding-wise hierarchical adaptation control. For real-time adaptation, the proposed service binding makes use of a meta-object-based reflective system.
• The proposed platform provides intelligent servicecontext management using a supervised and reinforcement learning-based machine learning framework.
• We implemented and evaluated a lightweight prototype of the proposed computing model. It shows enhanced performance compared with uniform binding-based legacy computing models. • The proposed platform can be applied to environmental sensors, actuators, service-agent devices, and network devices as a core function of computing infrastructure to provide intelligent IoT services in the future. Furthermore, the proposed platform can create new business opportunities in cloud computing where both the service providers and the network operator take co-responsibility for the performance, reliability, and scalability of the services.
CONCLUSION :
This paper presented a hierarchical cloud computing model for context-aware IoT services. It supports nonuniform service binding, real-time service-binding adaptation, and intelligent service-context management. We implemented a lightweight prototype of the proposed computing model and confirmed that the proposed model offers enhanced performance in terms of system throughput as compared with legacy uniform binding based computing models. The proposed computing model can be deployed to all information technology consumer devices and network entities as a key infrastructure.