Mario Fanelli

A survey of context data distribution for mobile ubiquitous systems

The capacity to gather and timely deliver to the service level any relevant information that can characterize the service-provisioning environment, such as computing resources/capabilities, physical device location, user preferences, and time constraints, usually defined as context-awareness, is widely recognized as a core function for the development of modern ubiquitous and mobile systems. Much work has been done to enable context-awareness and to ease the diffusion of context-aware services; at the same time, several middleware solutions have been designed to transparently implement context management and provisioning in the mobile system. However, to the best of our knowledge, an in-depth analysis of the context data distribution, namely, the function in charge of distributing context data to interested entities, is still missing. Starting from the core assumption that only effective and efficient context data distribution can pave the way to the deployment of truly context-aware services, this article aims at putting together current research efforts to derive an original and holistic view of the existing literature. We present a unified architectural model and a new taxonomy for context data distribution by considering and comparing a large number of solutions. Finally, based on our analysis, we draw some of the research challenges still unsolved and identify some possible directions for future work.

A Stable Network-Aware VM Placement for Cloud Systems

Virtual Machine (VM) placement has to carefully consider the aggregated resource consumption of co-located VMs in order to obey service level agreements at lower possible cost. In this paper, we focus on satisfying the traffic demands of the VMs in addition to CPU and memory requirements. This is a much more complex problem both due to its quadratic nature (being the communication between a pair of VMs) and since it involves many factors beyond the physical host, like the network topologies and the routing scheme. Moreover, traffic patterns may vary over time and predicting the resulting effect on the actual available bandwidth between hosts within the data center is extremely difficult. We address this problem by trying to allocate a placement that not only satisfies the predicted communication demand but is also resilient to demand time-variations. This gives rise to a new optimization problem that we call the Min Cut Ratio-aware VM Placement (MCRVMP). The general MCRVMP problem is NP-Hard, hence, we introduce several heuristics to solve it in reasonable time. We present extensive experimental results, associated with both placement computation and run-time performance under time-varying traffic demands, to show that our heuristics provide good results (compared to the optimal solution) for medium size data centers.

VM consolidation: A real case based on OpenStack Cloud

In recent years, Cloud computing has been emerging as the next big revolution in both computer networks and Web provisioning. Because of raised expectations, several vendors, such as Amazon and IBM, started designing, developing, and deploying Cloud solutions to optimize the usage of their own data centers, and some open-source solutions are also underway, such as Eucalyptus and OpenStack. Cloud architectures exploit virtualization techniques to provision multiple Virtual Machines (VMs) on the same physical host, so as to efficiently use available resources, for instance, to consolidate VMs in the minimal number of physical servers to reduce the runtime power consumption. VM consolidation has to carefully consider the aggregated resource consumption of co-located VMs, in order to avoid performance reductions and Service Level Agreement (SLA) violations. While various works have already treated the VM consolidation problem from a theoretical perspective, this paper focuses on it from a more practical viewpoint, with specific attention on the consolidation aspects related to power, CPU, and networking resource sharing. Moreover, the paper proposes a Cloud management platform to optimize VM consolidation along three main dimensions, namely power consumption, host resources, and networking. Reported experimental results point out that interferences between co-located VMs have to be carefully considered to avoid placement solutions that, although being feasible from a more theoretical viewpoint, cannot ensure VM provisioning with SLA guarantees. ? We discuss VM consolidation issues in Cloud Infrastructure as a Service (IaaS). ? We survey related works to clarify current state-of-the-art and ongoing research. ? We propose a management infrastructure for the open-source OpenStack Cloud. ? We highlight interferences due to network virtualization between co-located VMs.

Adaptive context data distribution with guaranteed quality for mobile environments

The capability to gather and timely deliver any relevant information useful in service-provisioning environments, typically called context data, is increasingly recognized as a core function of modern ubiquitous and mobile systems. However, notwithstanding the large amount of literature on context modeling and representation, much work is to be done in obtaining context data distribution and retrieval with guaranteed quality levels. This paper offers three main contributions along that direction. First, it proposes a simple way to specify context data distribution requirements: these objectives called context data distribution level agreement (CDDLA) focus mainly differentiated priorities, quality metrics, and delivery delay. Second, it presents how to translate CDDLA into a set of system-level parameters to drive the infrastructure configuration. Third, it describes our novel infrastructure for quality-aware context data distribution in mobile and wireless wide-area environments. The paper reports some experimental results to point out how our solution can grant context data dissemination with required quality, while achieving overall system scalability.

Increasing Cloud power efficiency through consolidation techniques

In the recent years, Cloud computing is emerging as the next big revolution of both computer networks and web provisioning. Due to its enormous promises, several vendors, such as Amazon and IBM, started designing, developing, and deploying Cloud solutions to optimize the usage of their own data centers. Unfortunately, several management issues of the Cloud are still open and deserve additional research. Among them, and fuelled by the emerging Green Computing research, Cloud architectures have to consolidate virtual machines in the minimal number of physical servers to reduce the run-time power consumption. In this paper, we present a project on power saving through server consolidation conducted at the IBM Innovation Centre in Dublin. Our experimental results, collected on a real testbed, show that server consolidation can effectively save energy, while introducing minimum performance degradation.

Implementing a scalable context-aware middleware

Recent advances in portable client devices are enabling new scenarios where mobile users assume to continuously access to various services, such as email, printing, and social computing applications, by opportunistically exploiting any computing resource and any wireless connectivity possibility encountered during their roam. That requirement calls for suitable context-aware middlewares capable of retrieving and using context data to personalize service provisioning. However, existing solutions still exhibit very limited support for context data management and distribution, especially in wide-scale and highly heterogeneous systems. The paper proposes a novel context-aware middleware to achieve scalability in context data dissemination. The primary design guideline is to reduce context data traffic by using a hierarchical distributed architecture and pervasive physical/logical caching techniques. To validate our design choices, we exploited our context-aware middleware to realize an advertisements service and a service discovery facility in our university campus. Obtained performance results demonstrate that our solution positively affects system scalability and average context dissemination time.

Self-Adaptive Context Data Distribution with Quality Guarantees in Mobile P2P Networks

Recent advances in wireless technologies and mobile computing are opening brand new opportunities for context-aware services, namely services able to tailor their behavior according to current execution context. Starting from the core assumption that only effective and efficient context distribution can pave the way to the deployment of truly context-aware services, this paper proposes a novel Peer-to-Peer (P2P) model and distributed architecture for context distribution in densely populated mobile systems. The primary guideline is to exploit agreed quality parameters on context distribution (mainly, delivery time and context freshness) and system-level monitoring information dynamically gathered, to take proper management decisions able to foster high scalability and low additional overhead. Our solution monitors both its own local components status and distributed run-time conditions to introduce two main self-adaptive capabilities, namely self-configuration and self-optimization. While the former lets our system automatically select data sending rates with no user intervention, the latter allows to reach better run-time performance by dynamically adjusting system components. The paper reports also several significant results, based on both a simulation-based implementation and a real-world prototype, showing that our solution can grant timely and efficient context delivery and dynamically achieves agreed quality levels with good quality-overhead tradeoff.

Towards Adaptive and Scalable Context Aware Middleware

The diffusion of portable client devices is promoting the spreading of novel mobile services, both traditional such as email and printing, and new such as social computing applications, capable of opportunistically exploiting any computing resource and any wireless connectivity encountered by roaming users. The new requirements call for novel context-aware middlewares to support and simplify the retrieval and the usage of context data. However, existing context data dissemination infrastructures still present several limitations: they are unable to adaptively exploit impromptu any wireless communication opportunity; they are unable to scale, especially in wide/densely populated environments; and they are prone to connection/device flaws. The article proposes a novel context-aware middleware that achieves adaptability, scalability, and dependability in context data dissemination through three main core guidelines: by using a distributed hierarchical architecture, by employing lightweight and adaptive context data dissemination solutions, and by adopting statistical context data/query replication techniques. The performance results, obtained by extensively testing the proposed solution in our wireless university campus testbed, have validated our design choices.

QoC-Based Context Data Caching for Disaster Area Scenarios

Disaster area scenarios are the consequence of sudden and unexpected disasters due to either human or natural causes, such as terrorist attacks and earthquakes. Towards the main goal of saving as many human lives as possible, context-aware services are emerging as standard-de-facto solutions to improve the coordination of involved rescue teams. Unfortunately, the scarce resources (communication bandwidth, memory, etc.) lead to very tough deployment scenarios to deal with. This paper presents our real-world quality-based context data distribution infrastructure for context-aware services in disaster areas. The main contribution is a quality-based caching solution that self-adapts by using quality requirements associated to close neighbors. Collected experimental results confirm that our proposal improves both system scalability and data availability.

Self-adaptive and time-constrained data distribution paths for emergency response scenarios

Emergency response scenarios where team members coordinate impromptu towards the common human rescue goal are pushing to the extreme the demand for novel services fully aware of any (dynamically collected) relevant information describing the disaster area, namely context-aware services. Unfortunately, the real-world realization of emergency response context-aware services poses several and still unsolved issues. Timeliness is crucial when dealing with safe critical data, while efficiency and reliability are necessary to guarantee services provisioning in disaster areas. This paper proposes an original solution to increase context data distribution scalability and reliability with agreed quality levels, mainly focusing on data retrieval time. The primary design guideline is to monitor and self-adapt the data distribution task by dynamically reorganizing (a limited number of) data distribution paths. The reported experimental simulation results point out that our solution can significantly reduce exchanged message number and fulfill agreed quality levels.