Computer Laboratory

  

LCDNet

To start (in 2015)

(1) Architecting the Internet for the Challenged (EU H2020 RIFE and UMOBILE projects)

In these projects we propose to address the challenges of making the Internet ubiquitous, accessible and energy-efficient. We do this by traversing a range of connectivity options that ensure universal coverage, while providing a single unifying communication architecture with a single set of abstractions that not only spurs innovation for a wide range of new services and applications but also encompasses existing successful Internet services. We utilise advances in information-centric networking (ICN) to provide this abstraction - an abstraction driven by access to and provisioning of information rather than the connection to explicitly identified endpoints.Through this abstraction, we accommodate today’s web-based services while providing a path to future immersive and sensor-rich applications, such as those envisioned by the Internet of Things. The focus on information enables inherent support for rich caching policies that ultimately increase the efficiency of the network across different technologies by providing information from the most efficient provider rather than the original source.

The concept of overarching ICN enables us to pursue multiple complementary connectivity options, specifically including Delay Tolerant Networking (DTN), as distinct dissemination strategies, each of which constitutes a set of protocols that optimally utilise local resources. Integrating multiple concurrent dissemination strategies enables the utilisation of connected and disconnected modes of access under a single architectural (information-centric) abstraction. This enables us to accommodate a pure IP-based world as much as a challenged connectivity DTN world, all within a single architecture, while exploiting all possible communication opportunities that particular network deployments provide, ranging from fixed, all-optical deployments of wireless and mobile networks to satellite-supported deployments that cover difficult-to-reach environments

In these projects we propose to address the challenges of making the Internet ubiquitous, accessible and energy-efficient. We do this by traversing a range of connectivity options that ensure universal coverage, while providing a single unifying communication architecture with a single set of abstractions that not only spurs innovation for a wide range of new services and applications but also encompasses existing successful Internet services. We utilise advances in information-centric networking (ICN) to provide this abstraction - an abstraction driven by access to and provisioning of information rather than the connection to explicitly identified endpoints.

Ongoing Projects

(1) Virtual Public Networks (EU FP7 COSMOS)

There have been several initiatives to enable wider access to the Internet. The Public Access WiFi Service (PAWS) is one such initiative that enables free Internet access to all and is based on Lowest Cost Denominator Networking (LCDNet) – a set of network techniques that enable users to share their home broadband network with the public. LCDNet makes use of the available unused capacity in home broadband networks and allows Less-than-Best Effort (LBE) access to these resources. LCDNet also enables third party stakeholders, such as local government, to become virtual network operators, reducing the costs of network operators to setup and manage new infrastructures to extend access to their Internet backhaul. With the advent of software defined networking (SDN), there are more opportunities for network operators to create, deploy and manage such open home networks at large scale. As such, we envision Virtual Public Networks (VPuN), i.e., home networks created, deployed and managed through an evolutionary SDN control abstraction. This offers more flexibility to users and network operators, allowing them to share and control the network, while providing opportunities for new stakeholders to emerge as virtual network operators.

VPuN are envisaged to expose a dynamic and user-friendly abstraction to stakeholders at various levels of the network to specify network resource requirements. When explicit requirements are not specified or if external factors compel it, the network must be able to reprogram itself automatically. This can be useful in emergency situations where access points can automatically mesh with access points of other users or personal devices based on online social networking trends (e.g., earthquake or tsunami in the area). Furthermore, network operators will be able to dynamically allocate capacity based on real-time needs (e.g., to support sudden increase in traffic due to flash crowds).


Past Projects

(1) Delay Tolerant Scavenger Transport for the Developing World

Internet access is crucial. However, more than 5 billion people are without Internet access. Although there have been several initiatives in the past trying to tackle this problem, the notion of enabling the other 5 billion have gained prevalence recently with the emergence of the Alliance for Affordable Internet, Internet.org etc. There are several ways of tacking the problem of solving access challenges: architecting new longer range low cost wireless infrastructures (see work done at the TIER berkeley), satellite access (see O3b) etc. There have been alternate approaches like the Public Access Wifi Service (PAWS) which uses wifi crowdsharing where existing Internet users share their home broadband connections with the poor for free. Even though Internet access infrastructures can be set up, backhaul Internet capacity is always a costly resource especially for areas with deprived connectivity. Recently Facebook's founder Zuckerberg has been pressing for compression of transmitted web data to reduce costly data usage.

The ubiquitous nature of cloud-centric apps and user generated content impose a serious challenge to the under privileged population who have limited access to costly Internet backhaul capacity. Such content (which do not have any strict real time requirements) consume a significant amount of the costly (and hence precious) Internet backhaul capacity. The project aims to tackle the problem imposed by such applications  enabling them to use scavenger transport methods
instead of the traditional TCP transport methods. Such methods will enable the apps to be transmitted without impacting other competing (real-time and hence more important) flows, efficiently utilising the network capacity. We also add an extra feature of delay tolerance within edge/access networks, where such non real time less priority applications can be transmitted using a store and forward approach and at the same time scavenging for any available capacity without
hampering other competing flows. Enabling delay tolerance to such apps will also be useful under the presence of unreliable wireless networks that witness frequent disconnections.

(2) Public Access WiFi Service

PAWS: Public Access WiFi Service project aims to utilise the unused capacity at home broadband networks to provide free Internet access to all to access essential services. The project extends the stakeholder value chain for incentivising Internet access deployment by including more than the two traditional parties (consumer and Internet service provider), for example, adding local government, who have a vested interest in decreasing the cost of human-centered service, and replacing these costly interactions with online services, which already prove popular with existing Internet users. This project is funded by the EPSRC. The project has several partners: Horizon Digital Economy Research Institute, Rural Digital Economy Hub (funded through its partnership fund), BT, SamKnows, Georgia Tech, Nottingham City Council and the Highlands and Islands Enterprise.


(3) New Satellite Access Methods to Provide Lower than Best Effort Internet Access

This is a project funded by the French Space Agency and Thales Alenia Space to explore new low-cost satellite access methods. The project is in collaboration with TESA/ISAE.