Sustainable Smart Computing

Click on the links below for more details.

Cloud Computing

Synchromedia leads pilot research projects developing various cloud computing platforms driving state-of-the-art next generation data centers and networks. Cloud computing links all resources together in a common architecture that can be virtualized. In other words, the compute and storage platform is architecturally "unified" with the network and the virtualization platform, making no distinction between the network and the edge devices connected to it. Cloud computing eliminates manual integration, in favor of an integrated architecture, thereby breaking down the silos of compute, storage, and network resources. Sample projects carried out by Synchromedia in this field include Synchromedia Platform (CFI), GreenStar Network (CANARIE), and Green Sustainable Telco Cloud (Ericsson).

 

Network Virtualization

Network virtualization research at Synchromedia is aimed at creating new models providing a logical software-based view of the underlying hardware and software networking resources (switches, routers, etc.). This results in an intelligent abstraction that makes it easy to deploy and manage network services and resources. We are pioneering the NFV (Network Function Virtualization) concept and very active in SDN (Software-Defined Networking) solutions, especially with respect to intra and inter data center networks. Sample projects carried out by Synchromedia in this field include UCLP (CANARIE), PanLab 1 (EU’s FP6 Framework), PabLab 2 (EU’s FP7 Framework), Mantychore (EU’s FP7 Framework), GreenStar Network (CANARIE) and V-WAN hypervisor (Ciena).

Sustainable and Environment Awareness

Relying heavily on the computing power of data centers, the Internet is enabling the establishment of entirely new industries and unlocking vast potential for innovation in existing ones, like education, health care, energy, the public sectors, and knowledge dissemination. Advanced countries, like the US, have proposed national plans to ensure that every citizen will have access to ultra-broadband Internet. As a result, larger data centers are required to host increasingly critical applications. This raises new questions, in terms of scalability, resiliency and operational cost-effectiveness, and power consumption and environmental concerns in particular. Novel paradigms are therefore required to reduce massive waste due to comatose servers and overprovision, as well as new energy efficient approaches to design and manage data centers. Synchromedia is worldwide recognized as leader in Green ICT, Sustainability and Environmental Awareness research. In collaboration with the Canadian Standards Association (CSA), Synchromedia introduced the first Canadian standard for green ICT projects in 2011. It is also an active member of ITU and many other research institutes on life-cycle assessment for ICT.

Network and Cloud Security

The potentially revolutionary cloud computing and networking paradigms could not be achieved in reality without appropriate security and privacy solutions. Designing trustworthy clouds or interoperable secure clouds is very challenging, because the heterogeneity and diversity of the services provided, as well as the diverse access requirements of domains in cloud environments, demand fine-grained access control policies. Synchromedia developed new security services for cloud platforms by integrating privacy-preserving protocols to verify various identity attributes through zero knowledge, proof-based techniques. The framework also allows to efficiently capture a generic set of the parameters that are essential to establishing trust and to managing evolving trust and interaction/sharing requirements. Our research addresses semantic heterogeneity, secure interoperability and policy evolution management, as well as reliability violations.

Traffic Modeling and Big Data analytics

Synchromedia research has a strong focus on smart traffic analyzer and QoS provisioning based on big data mining. Various models have been built for both intra and inter- data center network, and a framework for optimally allocating resource according to each traffic model has been developed, in particular with respect to Telco and HPC applications. Based on advanced learning algorithms, relationship between application requirements and underlying traffic is discovered. These patents are contextualized to understand the behaviours of network elements, applications, and data centres, thus help optimize operational activities. Traffic modeling and big data analytical methods introduced by Synchromedia team have effectively been used in both cloud computing and Future Internet fields.

Wireless Communication

Synchromedia has recently conducted large-scale research projects on wireless communications in collaboration with Ericsson. A cloud-based platform has been developed to support virtual wireless communications across heterogeneous access networks and extent mobile content distribution across cloud and mobile devices. The research framework covers various types of wireless technologies, such as WiFi, ZigBee, LTE, and 5G, as well as ubiquitous multiple dimensional communications between wireless and optical network. Synchromedia research outcome is used to provide next-generation wireless communications to ETS campus and Montreal’s Griffintown district.

Behavioral Learning

Novel cognitive sciences and machine learning methodologies are developed by Synchromedia team dealing with efficient management in cloud computing environment. Cognitive radio network models have also been proposed for dynamically selecting transmission bands and routing paths, helping to maximize network throughput by performing joint routing, dynamic spectrum allocation, scheduling and power transmission control. Cognitive techniques allow mobile users to select a nearby data centre, thereby reducing delays and jitter. In particular, cognitive orchestration solutions have been introduced to forecast Ericsson’s Telco cloud activities, taking into account environmental impacts and end-user quality of experience (QoE). Advanced statistical machine learning (SML) methods have been proposed to build diagnostic and predictive tool that allow dynamic scaling, automatic reaction to performance and correctness problems, and generally automatic management of many aspects of data center operation.
ericssonlogo
inocybelogo
canalogo
cienalogo
Civimetrix Telecom logo
mitacslogo
risq logo
nserclogo
promptlogo
ecolepolytechniquelogo
University of Torontologo
frqntlogo
uqlogo
MDEIE logo
cfilogo
ciraiglogo