Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Intellileaf: A Novel Framework Enabling Intelligence on Resource Constrained Internet-of-things
Communication technology has led to broader and continued connectivity between humans ever since the discovery of radio waves. The concept of connectivity has been extended from connecting humans to connecting machines in recent times. The Internet-of-Things (IoT) paradigm has led to an exponential increase in the number of devices, which will potentially be connected to the broader web. The coming of 5G networking capacity harks the densification of networks, increase in heterogeneity of devices on the network and increase in the data being exchanged on the network. Machine intelligence is a parallel domain of development, which aims to enable machines to make decisions based on prior event-based learning or situational awareness. Machine intelligence promises safety aware industrial robots, autonomous factory floor navigation, self-configuring robotic swarms achieving a single task, among a myriad of other applications. 5G supplements this domain by allowing higher than ever bandwidth for devices to exchange information. As of today, the focus is mainly on developing edge networks. These are networks where the business logic or decision-making rules or processing power rests at the resource unconstrained devices at the edge. The ‘edge’ constitutes the layer of devices right above the end-devices in the network hierarchy. This allows for lower round-trip-times (RTT) decreasing communication latency. The densification of the end-device network in the IoT paradigm enables observation at a larger scale. This densification allows for general observations specific to the dense end-device network, which can be leveraged to utilize or route information intelligently. In this work, a framework is presented, which abstracts the finer details of a complex network to a collection of essential actions and leverages the group behavioral observation of dense end-device networks to enable efficient network usage. Previously, the use of sophisticated intelligence algorithm on resource-constrained end-devices was not possible due to limited processing and energy resources. The use of group observation in this framework allows the implementation of intelligence on resource-constrained end-devices leading to better overall network performance at minimal energy cost per network end-device. This will extend the advantages of fast, intelligent networking to the end-device layer. The abstraction of the network to essential building blocks will allow faster and tractable network modeling and simulations foregoing the need for complex deployment specific modeling efforts or deployment of physical network infrastructure.
Rakshit, Sushanta Mohan, "Intellileaf: A Novel Framework Enabling Intelligence on Resource Constrained Internet-of-things" (2019). ETD collection for University of Nebraska - Lincoln. AAI13862686.