Wireless Communication Protocols for Distributed Computing Environments

TL;DR: The aim of this chapter is to present the most important challenges for the communication point of view when forming a distributed network for performing parallel and distributed computing by considering their capabilities in terms of reliability and adaptation when facing with heterogeneous computing requests.

Abstract: The distributed computing is an approach relying on the presence of multiple devices that can interact among them in order to perform a pervasive and parallel computing. This chapter deals with the communication protocol aiming to be used in a distributed computing scenario; in particular the considered computing infrastructure is composed by elements (nodes) able to consider specific application requests for the implementation of a service in a distributed manner according to the pervasive grid computing principle (Priol & Vanneschi, 2008; Vanneschi & Veraldi, 2007). In the classical grid computing paradigm, the processing nodes are high performance computers or multicore workstations, usually organized in clusters and interconnected through broadband wired communication networks with small delay (e.g., fiber optic, DSL lines). The pervasive grid computing paradigm overcomes these limitations allowing the development of distributed applications that can perform parallel computations using heterogeneous devices interconnected by different types of communication technologies. In this way, we can resort to a computing environment composed by fixed ormobile devices (e.g., smartphones, PDAs, laptops) interconnected through broadband wireless or wired networks where the devices are able to take part to a grid computing process. Suitable techniques for the pervasive grid computing should be able to discover and organize heterogeneous resources, to allow scaling an application according to the computing power, and to guarantee specific QoS profiles (Darby III & Tzeng, 2010; Roy & Das, 2009). In particular, aim of this chapter is to present the most important challenges for the communication point of view when forming a distributed network for performing parallel and distributed computing. The focus will be mainly on the resource discovery and computation scheduling on wireless not infrastructured networks by considering their capabilities in terms of reliability and adaptation when facing with heterogeneous computing requests.