Toward Verifiable Services: Bound-Oriented Network Tomography
DescriptionNetwork tomography has been an important research direction in the past decade. Roughly speaking, network tomography is to infer certain network performance that is difficult or too costly, to be measured directly. The inference is derived based on a limited number of measurements that are partially or indirectly related to the target to be inferred. In the past, people have studied network tomography for different purposes, such as inferring network topology and delay of links.In this project, we study a class of new network tomography problems, referred to as bound-oriented network tomography. The problems are particularly relevant to ISPs who need to continuously invigilate the status of network resources important to their operations, for example, when the ISPs are providing verifiable services to their clients. Effective and efficient network tomography is a basic problem faced by them.Concerning the need from ISPs, our problems have two important features that have been overlooked in the past. First, we focus on inferring performance bounds rather than the absolute values. The bound-oriented inference is consistent with the fact that the SLA terms between ISPs and their clients are commonly written in terms of performance bounds. Second, for each path to be inferred, we directly infer the performance bounds of the path without always explicitly obtaining the performance of each link on the path. Knowing path performance is enough to ISPs because their service levels to clients depend on paths directly. These two features lead to new models that are less restrictive and enable low-cost solutions.Regarding the plan and methodologies, we propose three concrete tasks, inferring delay bounds, inferring bandwidth bounds, and inference by network calculus. Among the three, the first two are independent, both needing some measured bounds as input. The third task will provide valid input to the first two tasks; at the same time, the third task will also infer more information by a new concept termed as service curve decomposition. Successfully doing service curve decomposition will expand the applicable area of network calculus, making it a powerful tool for network tomography.In summary, we propose a framework for bound-oriented network tomography, which can be used for ISPs to fast gain network knowledge at low cost.?
|Effective start/end date||1/01/18 → …|