Mobile Micro-Cloud: Application Classification, Mapping, and Deployment
Shiqiang Wang^, Guan-Hua Tu#, Raghu Ganti*, Ting He*, Kin Leung^, Howard Tripp+, Katy Warr+, and Murtaza Zafer* # UCLA – US, * IBM – US, ^ Imperial College – UK, + Roke Manor – UK
Abstract— Mobile micro-cloud envisions a logical network composed of two components, the core (e.g., the command and control center) - with access to large quantities of static (and possibly stale) information and the edge (e.g., the forward operating base) - with access to smaller quantities of more real-time and dynamic data. The edge and core are separated by dynamic and performance constrained networks with a many-to-one relationship between the core and the edge. The goal of the mobile micro-cloud is to deliver situational awareness to the small units (primarily interacting with the edge) in a timely and resource aware manner. Fundamental to this mobile micro-cloud paradigm is the flexibility for users to deploy varied applications dynamically as demands, capacity, connectivity, and mission requirements continuously evolve. This “runtime” approach is in contrast to historical systems that are provisioned based on fixed requirements for specific applications. In this paper, we examine various aspects of the mobile micro-cloud. First, we present an approach to deriving semantics for consistent representation of application requirements in order to enable a generic approach to application deployment in the mobile micro-cloud environment. Second, we examine the advantages of migrating an application (or service) to the edge and quantify these gains through preliminary experimental results. Third, we examine the problem of mapping applications (identified for migration) to available resources that are changing dynamically in a Security-aware manner. Finally, we illustrate the prototype platform for the mobile micro-cloud and its characteristics. Keywords— Mobile micro-cloud, application placement
Mobile micro-cloud envisions that applications (or computing tasks) will be deployed in a mobile micro-cloud, a logical network composed of two components, the core (e.g., the command and control center) – with access to large quantities of static (and possibly stale) information and the edge (e.g., the forward operating base) – with access to smaller quantities of more real-time and dynamic data. The edge and core are separated by dynamic and performance constrained networks with a many-to-one relationship between the core and the edge. It is also possible for edge nodes to communicate with each other. Further, the (edge and core) nodes can belong to different coalition partners, raising the question of security and operational policies for handling of data and computation. Figure I-1 illustrates the vision of the micro-cloud for the delivery of situation awareness to the tactical edge and Figure I-2 illustrates a typical architecture of the mobile micro-cloud in the army coalition context.
Figure I-1: Using micro-clouds to deliver Situational Awareness to the tactical edge
Figure I-2: A tactical network scenario – Enabling efficient computations over dynamic networks
The benefits of embedding storage and computation into such a micro-cloud tactical network are two fold: (i) Effective provisioning for diverse information requirements – the micro-cloud supports users with different latency requirements and access rights and (ii) Effective information exchange in a constrained environment – Complete shuffling of information is impractical in a tactical network and the micro-cloud reduces congestion by providing computation at the edge.
A natural question that arises is that of the applicability of traditional cloud computing technologies to the dynamic tactical network based micro-cloud environment. Traditional cloud computing has become quite popular in the recent years, providing utility computing in a flexible, agile, and scalable...
References:  M. Armburst et. al., “Above the Clouds: A Berkeley View of Cloud Computing”, Technial Report UCB/EECS-2009-28, UC-Berkeley, 2009.
 A. Fischer, J. Botero, M. Beck, H. De Meer, and X. Hesselbach, “Virtual network embedding: A survey,” IEEE Commun. Surveys Tuts., Feb. 2013, accepted.
 P. Bahl, R. Han, L. Li, and M. Satyanarayanan, “Advancing the State of Mobile Cloud Computing”, In Proceedings of MobiSys workshops, MCS, 2012.
 H. Cavilla, J. Whitney, A. Scannell, P. Patchin, S. Rumble, E. Lara, M. Brudno, and M. Satyanarayana, “SnowFlock: Rapid Virtual Machine Cloning for Cloud Computing”, In Proceedings of EuroSys, 2009.
 M. Chowdhury, M. Rahman, and R. Boutaba, “Vineyard: Virtual network embedding algorithms with coordinated node and link mapping,” IEEE/ACM Trans. Netw., vol. 20, no. 1, pp. 206–219, 2012.
Please join StudyMode to read the full document