Majority of software-defined network design and implementation efforts have been focused on local area networks. These environments are highly controlled, with reasonable control over network failures, and very few truly random event occurrences. SD-WANs extend the traditional notion of SDNs to the wide area network where SDNs will face three fundamental challenges: increased latency, low-average bandwidth, and an increased probability of failures. We are currently exploring the design of distributed scalable, fault-tolerant and secure SD-WANs. Availability and tolerance towards network partitions have been well-documented, but their impact on consistency in moderate-bandwidth, increased latency environments is not well studied. We aim to architect SD-WANs using a new paradigm of Distributed Network Services (DNSs) as building blocks. These DNSs are extensions of ONUG’s vision of Network Service Virtualization (NSV), where service capabilities are tied to each instance of an application and are orchestrated and instantiated with the application.
Rural areas continue to be plagued by limited data connectivity due to poor and unreliable power infrastructure, poor backhaul connectivity and lack of economic incentives for telecom providers. These factors continue to be an impediment to providing reliable, highly available and bandwidth intensive mobile services and applications to function well in such contexts. This paper presents the design, implementation and deployment of GreenLinks, a new ground-up platform that enables intermittency-aware, reliable and available cellular services and application support in rural contexts under extreme operational environments with limited power and no existing cellular coverage. Unlike the conventional monolithic cellular network architecture, GreenLinks enables a distributed (and potentially disjoint) collection of open, programmable cellular base stations to offer a gamut of both conventional cellular services and new forms of distributed edge services. Using a participatory sensing white spaces approach, a GreenLinks base station can co-exist with other conventional cellular networks.
Modern web pages are very complex; each web page consists of hundreds of objects that are linked from various servers all over the world. While mechanisms such as caching reduce the overall number of end-to-end requests saving bandwidth and loading time, there is still a large portion of content that is re-fetched – despite snot having changed. In this demo, we present Extreme Cache, a web caching architecture that enhances the web browsing experience through a smart pre-fetching engine. Our extreme cache tries to predict the rate of change of web page objects to bring cacheable content closer to the user.
Conventional routing protocols have been designed with the objective of selecting the best paths in a network based on a defined routing metric. We introduce Maximum Entropy (Max-Ent) Routing, a new class of routing protocols with the design philosophy of explicitly relaxing the notion of best paths to increase the freedom of choice of available paths to each destination. In this paper, we demonstrate how the Max-Ent routing framework can be applied in the context of distance vector and path-vector routing protocols. While Max-Ent routing may choose sub-optimal paths, we explicitly show how to modify the route selection metric in both these vector protocols to achieve the dual properties of maximum entropy and route convergence. Using the RCP framework of having a central controller running the Max-Ent route manager for an ISP, we show how a single AS can benefit by using the Max-Ent routing framework to minimize the routing damage experienced at a node in the face of node or edge failures.