Abstract of Talk
Distributed Services in Pervasive System
Abstract: Pervasive systems have found applications in various areas such as entertainment industry, healthcare, military and others. Devices in a pervasive system are generally resource constrained, heterogeneous, personal, and mobile. Each device may offer a set of software services to other devices in the network. A pervasive device might need to perform data processing tasks such as video and audio filtering in such applications as monitoring of elderly in homes for the aged. Devices may need to perform such tasks even if the required software is not installed locally on the device. Since the devices are resource constrained, a device might not have sufficient residual energy to complete the task. Hence collaboration between devices is exploited to perform computational tasks. In the past, service discovery, service composition and cyberforaging techniques have been used to facilitate resource sharing amongst devices in pervasive systems. Some researchers have performed offloading of computationally intense tasks from resource constrained mobile devices to high end servers. These servers are used to perform decision making and task execution. Such high end and centralized servers might not be available within a pervasive environment. Study of decentralized, autonomous and distributed execution of the tasks in pervasive systems is available in a very limited amount in the literature. This dissertation provides algorithms for cooperative service executions in pervasive environments that might not have high end or centralized servers. Pervasive systems are classified into smart spaces, Mobile Ad hoc Networks (MANETs), Vehicular Ad hoc Networks (VANETs) and Opportunistic Networks (OPNETs). The major contributions of this dissertation are: • Cluster Based Scheduling (CBS) algorithm that performs service scheduling in smart spaces. • Decentralized Grading Autonomous Selection (DGAS) algorithm that performs fault tolerant service execution in Mobile Ad hoc Networks. • Mutual Exclusion for Opportunistic Networks (MEOP) algorithm that facilitates exclusive access to shared resources in opportunistic networks. • An algorithm for performing service composition in opportunistic networks. The main advantages of CBS over existing schemes are: reduced communication and storage overhead, and support for usage of multiple task scheduling algorithms. CBS factors such challenges as device heterogeneity, service availability and device mobility into the scheduling algorithm. CBS schedules tasks that have dependencies amongst each other. DGAS assigns independent tasks onto mobile service providers. Service execution on devices in pervasive systems might suffer from node and link failures. DGAS provides fault tolerance by replicating service execution onto multiple devices. When multiple devices need exclusive access to a shared resource, the middleware installed on the devices should facilitate the access. MEOP is a token requesting algorithm that provides exclusive access to resources in opportunistic networks. MEOP has lower communication overhead as compared to token ring algorithms. MEOP can be used along with any routing protocol. This dissertation also presents service composition algorithm for opportunistic network. Analysis of success probability of task execution and the length of compositions is presented and verified. A prototype of the service composition algorithms is implemented on PDAs and netbooks.