CSR---VCM: Managing a Concurrent Opportunistic Sensor Environment with Pocket Hypervisors

并发机会主义传感器环境(COPSE)旨在创建一个分散的对等平台，以利用地理区域内的计算资源来提供移动和普及服务。区域内的节点可以是固定元素(如接入点)和高移动性设备(如移动电话)的混合。节点自愿提供诸如虚拟机托管或互联网连接等资源，并且可以在不通知的情况下撤销资源。服务跨一组不同的元素执行，并由单个虚拟机实例化组成。根据运行它们的节点的不同，服务实例化可能会跨不同的能源预算、通信技术和存储配额执行。为了让Copse取得成功，它必须为服务组织足够的自愿资源，以便服务运行，并公开有关快速变化的资源可用性的信息，以便服务能够适应人员流失。这需要在两个大的领域应对研究挑战。第一个区域是虚拟机接口，它公开本地资源和共存节点中的更改；这些接口将允许服务重新配置自身，以响应资源流失。第二个领域是将参与COPSE等协作移动系统的机会成本降至最低；袖珍管理程序必须提供保护设备所有者和同时运行的服务免受错误或恶意实例化影响的机制和策略。

Concurrent opportunistic sensor environments (COPSE) aim to create a decentralized peer-to-peer platform for harnessing the computational resources within a geographic area for mobile and pervasive services. The nodes within an area can be a mix of fixed elements like access points and highly mobile devices like mobile phones. Nodes contribute resources such as virtual machine hosting or Internet connectivity voluntarily and can revoke resources without notification. Services execute across a heterogeneous set of elements and are composed of individual virtual machine instantiations. Depending on the nodes running them, service instantiations may execute across different energy budgets, communication technologies, and storage quotas. For COPSE to succeed, it must marshal enough volunteered resources for services to run and expose information about the rapidly changing availability of resources so that services can adapt to churn. This requires meeting research challenges in two broad areas. The first area is virtual machine interfaces that expose changes in local resources and co-located nodes; these interfaces will allow services to reconfigure themselves in response to resource churn. The second area is minimizing the opportunity costs of participating in cooperative mobile systems like COPSE; pocket hypervisors must provide mechanisms and policies that protect device owners and concurrently running services from buggy or malicious instantiations.