Research at INRG covers all levels of networking design, including data center design, network security, distributed sensor networks for a variety of applications (including healthcare and wildfire monitoring), machine learning algorithms for a variety of applications (including data center traffic load balancing and improving network performance), and more.
IoT Network Deployment Over 2.5D Terrain
One can use an outdoor, Internet of Things sensor network to detect events visually, but there must be clear line-of-sight (CLOS) between a sensor and the location of an event. This research project explores various deployment algorithms for finding locations for a network of IoT nodes to maximize the chance of detecting events.
PANTHER
Destructive wildfires have recently caused catastrophic and unprecedented damages in California and different parts of the world. Despite significant fire prevention efforts, wildfires are anticipated to grow in frequency and intensity due to the changing climate and shifting urbanization patterns. This research project aims to develop and deploy a large-scale IoT network consisting of miniature weather stations capable of monitoring various environmental conditions (e.g., temperature and humidity, wind, soil moisture) to help predict and detect wildfires in a timely manner and mitigate their potentially devastating consequences.
PIMAP
Pressure injuries are localized damage to the skin and/or underlying tissue that usually occur
over a bony prominence. Pressure injuries most commonly develop in individuals who have
low-mobility, such as those who are bedridden or confined to a wheelchair and consequently are
attributed to some combination of pressure, friction, shear force, temperature, humidity, and
restriction of blood flow and are more prevalent in patients with
chronic health problems.
DEL: Decentralized Edge Intelligence
The proliferation of data generated at the network edge and the increasing networking and computational capabilities of edge devices has given rise to applications that harness data to provide services and has given prominence to developing better networking, storage, and computation infrastructures at the edge. These enabling technologies, artificial intelligence and edge computing, has manifested in a new field — edge intelligence. In this research, we aim to provide support to the growth of edge intelligence by proposing frameworks for fully decentralized learning to enable full-spectrum edge intelligence, developing simulation platforms to allow for network-cognizant distributed learning experiments, and more.
Adaptive, Dynamic Load Balancing of Datacenter and WAN flows
Recent years have seen a rapid growth in the number of cloud computing applications, connected IoT devices. With the large scale deployments of 5G in the near future, there will be even more applications, including more bulk transfers of videos and photos, augmented reality applications and virtual reality applications which take advantage of 5G’s low latency service. All these add to heavy, bulk of data being sent to the data centers and over the backbone network. These traffic have varying quality of service requirements, like low latency, high throughput and high definition video streaming.
This research project focuses on designing a load balancer for data center networks that is adaptive to the kind of traffic it encounters by learning from the network conditions, and providing low latency and high throughput performance with increased network utilization.
GDSim
As cloud providers scale up their data centers and distribute them around the world to meet demand, proposing new job schedulers that into account data center geographical distribution have been receiving considerable attention from the data center management research and practitioner community. However, testing and benchmarking new schedulers for geo-distributed data centers is complicated by the lack of a common, easily extensible experimental platform. To address this gap, we propose GDSim, an open-source job scheduling simulation environment for geo-distributed data centers that aims at facilitating development, testing, and evaluation of new geo-distributed schedulers.
Network Security based on Cross-Layer Device Fingerprinting
Device fingerprinting (DF) has emerged as a promising technique against impersonation or insider attacks in wireless networks, enabled by signal analysis. DF can considerably enhance device identification accuracy and robustness, thus significantly reducing the risk of device impersonation attacks by leveraging unique characteristics of devices at the physical and MAC layers, termed hereafter as Cross-Layer (CL)-device fingerprinting (DF). This project covers the following aspects: 1) cross-layer features extracted based on applications, such as security, management, or network performance; 2) machine learning techniques applied on the extracted features for certain applications, such as distributed learning, on-line learning, unsupervised learning; 3) security enhancements based on device fingerprinting; 4) applications scenarios, such as wifi networks, vehicular networks, etc.
ADEPT: Adaptive Decentralized Emergent-behavior based PlaTooning
Vehicles following each other in close-proximity in the form of a platoon to reduce fuel consumption due to reduction in air drag saving up to 5% for the first vehicle and 10% for the following vehicles. Introduction of vehicle-to-vehicle communication has led to further decrease in inter-vehicular spacing, enabling further reduction in fuel consumption and improving the road capacity by up to 166%. Dynamic platoon formation and dissolution is generally governed by the lead vehicle, leading to a centralized approach. This suffers from various drawbacks such as single point of failure and performance bottlenecks resulting in serialization of maneuvers and shorter platoon length. A decentralized approach mitigates these issues and is traditionally realized using the “top-down” approach by first identifying the maneuvers and then executing the required steps and exchanging the necessary messages. Such decentralized platooning systems, which are known as “Deliberate Platooning”, suffer from higher complexity and reduced flexibility. This dissertation presents ADEPT – Adaptive Decentralized Emergent-behavior based PlaTooning, that aims at mitigating the shortcomings of “traditional” decentralized platooning systems. ADEPT uses a novel decentralized automated vehicle platooning approach inspired by nature’s “emergent behavior” commonly found in biological systems. In ADEPT, each vehicle follows a set of simple rules when they need to interact in order to carry out the maneuvers. Platooning maneuvers such as join, exit, and merge “emerge” as a result of vehicles following these “emergent rules”. Through extensive simulation experiments using a platooning-enabled vehicular network simulator driven by a wide range of scenarios, we demonstrate that ADEPT yields superior performance in terms of maneuver time and communication overhead, especially in multi-vehicle multi-maneuver scenarios. Additionally, we devise mechanisms for obstacle avoidance, vehicle following, gap- and predecessor determination on curved roads and, using a well-known simulation environment equipped with a physics engine, demonstrate that our emergent platooning approach is effective when vehicle characteristics such as weight, center of mass and friction are considered.
(Anuj’s)
(Sinjoni’s)
(Chris’s)
Past Projects
Bus Tracking System Project
BTS is a grad student project developed by i-NRG and other students to provide real-time tracking of the campus shuttles.
COMMUNITY
This INRIA – UC Santa Cruz Team investigates a number of research challenges raised by message delivery in environments consisting of heterogeneous networks that may be subject to episodic connectivity.
DrIVE Associated Team
Enabling ITS through programmable networks
Mobility Modeling in Wireless Networks
In this project we study mobile wireless networks by looking at mobility management and analysis of human mobility, focusing on the main goal of understanding human mobility and applying our findings on developing new realistic mobility models for simulations.
Socially- and Geographically-Aware Modeling Framework for User Mobility in Wireless Networks
One of the project’s main goal is to develop novel, socially-inspired human mobility models that also account for geographic diversity of the region of interest. The proposed models consider features observed in real human networks such as differential popularity, transitivity and clustering as well as geographical features and preferences from users. Another important deliverable of our work is a suite of tools that implement the proposed models and that can be used by other researchers and practitioners in the evaluation of mobile systems and protocols.
TerrainLOS
TerrainLOS is a propagation model (a propagation model is an algorithm that simulates real world wireless communication) that uses terrain to determine the success of a transmission.