Communication Research Group
CRG1: SDN, NFV and Cloud: Enablers for Mobile Network Architectures
The success of network operators in the 5G arena will depend heavily on their ability to shift from their old operational models, and embrace new technologies, design principles and innovation in both the business and technical aspects of the environment. A comprehensive review of both these business and technical aspects is needed to facilitate a clear roadmap for what is needed. The high level objectives of this project are:
- Creation of a cloud computing playground that facilitates research in Software Defined Networks (SDN) as well as Network Function Virtualisation (NFV)
- Identify unique solutions for facilitating core network sharing solution for multiple network operators
- Address service orchestration issues for end users by designing a specialized framework that takes into account users’ contexts
- Solutions for mobility management and resource allocation in a network-shared NFV environment. While much work looks at core network sharing, this work is focused on the sharing of the network access portion. Network sharing does pose challenges in the regulatory space.
- The design and deployment of an IP spoofing detector that uses Operating System (OS) fingerprinting to determine the true source of an incoming packet during DDoS attack in cloud computing environments.
- To facilitate the paradigm shift from Cloud to Mobile Edge Computing: design and implement a smart pre-copy live migration that intelligently predicting the dirty rate of benchmark workload to determine and alter the current stop and copy conditions. To thereby reduce the total migration time and down time experienced during VM live migration in Xen server
- Investigate the circumstances under which SDN can meet the requirements for carrier-grade resiliency. A number of different recovery methods will be addressed, such as route protection and restoration, and controller replication and redundancy. Additionally, a scenario in which complete controller failure occurs will be investigated. Where current methods are insufficient, novel methods will be developed to improve resiliency. These methods will be investigated by simulating and/or emulating an SDN. Additionally, a comparative legacy network will also be simulated, in order to provide a reference case to measure the SDN performance against.
- A further aim of this project is to investigate the scalability of the resiliency solutions. Many current research works have focused on relatively small networks. A large-scale SDN will be simulated and/or emulated, in order to provide a more realistic indication of the resiliency of SDN in a carrier network.
CRG2: Efficient Radio Resource Management in Next Generation Networks & Cognitive Networks
Efficient radio resource management is a primary concern in next wireless network due to limited availability of radio resources.
The objectives are to:
- Develop packet scheduling mechanisms to enable multipath packet transmission for multi-homed terminals for enhanced throughput and QoS provisioning.
- Develop network-selection algorithms to improve connection-level and packet-level quality of service in next generation wireless networks by using intelligent resource allocation mechanisms.
- Use optimization and pricing algorithms to ensure cost effectiveness in resource allocation and reduce congestion during peak hours in next generation wireless networks.
- Develop distributed mobility management schemes for seamless, efficient, and scalable service delivery in next generation wireless networks.
- Develop spectrum handoff schemes for cognitive radio network.
- Design and implement dynamic spectrum reallocation schemes that use autonomic management for efficient spectrum utilization in next generation networks.
- Investigate the feasibility of using millimetre wave for small cell backhaul in urban area, and develop efficient and cost-effective backhaul solutions for high-density small cells in urban areas by applying sensing, decision, and learning concepts of cognitive radio technology.
- Develop efficient radio resource management algorithms to mitigate interference, ensure fairness in radio resource allocation and provide QoS guarantee for hybrid access mode of small cells (femtocells) in nest generation wireless network. The developed algorithms are expected to improve overall network throughput, and enhance transmit power efficiency and reduce packet loss.
- Network Coding in Multicast Wireless Networks one strives to optimize energy consumptions in wireless multicast networks studying existing methods of minimizing energy in wireless multicast networks and propose novel approaches that are based on Data Envelopment Analysis (DEA) method to further optimize energy consumption in wireless multicast networks.
In Software-Defined Radio (SDR) and Cognitive Networks (CRN) the immediate objectives are to:
- Develop A New MAC Protocol for Optimal Spectrum Identification, Co-ordination and Scheduling in an Ad hoc Cognitive Radio (CR) Network for optimised wireless operation
- Develop adaptive cross layer transmit antenna selection algorithms for underlay CR MIMO systems that improve the performance of the cognitive radio user while keeping the interference at the primary user below a predetermined threshold value.
- Design, simulate and experiment on protocols that can be used to manage the various types of CRNs, using distributed and cooperative spectrum sensing so that the interdependence amongst the CRNs is of mutual benefit. This symbiotic relationship should improve the spectral efficiency and performance of the CRNs.
- Design algorithms that will be energy efficient for interference management in heterogeneous networks. This is a major project which will be carried out in four phases.
CRG3: Future Internet Research and Experimentation – IoT/M2M
The main objective of this project is to perform research in the broad area of Future Internet Networks and Applications. With this goal in mind, we have been able to participate in the development of 2 such experimental testbed projects. One testbed allows for the development of applications and solutions for IoT and Smart Cities network requirements. The second testbed explores the development of future network architectures (4G/5G) based on the upcoming technology enablers such as SDN and NFV in a locally built cloud computing environment.
The specific objectives to be reached in the project can be summed up as follows:
- To investigate, develop and test network infrastructure deployment solutions for Future Internet Technologies, 4G/5G Mobile Network Architectures and Services;
- To gain competence in the emerging areas of Software Defined Networks and Network Function Virtualization, which relates to the implementation of network architectures and equipment in cloud computing environments;
- Building on the two above points, the project will focus on research and development of adequate solutions for relevant Smart City and IoT applications in the South African context. These range from the domains of Energy Management, Smart Home, Smart Energy, eHealth, and disaster management etc. The focus of these solutions will take into account the challenges faced in South African (and to a wider context, African) cities;
- Strengthen the interconnection and extension of existing experimental facilities across continental boundaries with a specific focus on Smart Cities and Smart Energy; Integration of software-based cross-industry horizontal M2M frameworks with real world sensors and IoT device deployments;
- Usage of autonomic communication methods for end-to-end M2M communication in Smart Cities focusing on smart energy management;
- Final implementation and integration of delay tolerant capabilities to support opportunistic information transmission; Leverage existing standards in the field of M2M (ETSI TC M2M, oneM2M, OMA WM2M) and foster their global adoption;
- Implementation and finalization of user application for environmental monitoring as well as user and administrator applications of the smart energy system;
- Deployment and federation of testbeds across countries and continents, with a focus on underdeveloped countries and structurally challenged areas. The complete experiment environment can be deployed with the jFed experimenter GUI;
CRG4: Smart Cities
High level project objectives:
- To design, develop and test an open and cost-effective architecture for smart cities deployments and applications in Africa that should ease the development and implementation of locally designed and customized applications.
- To assist on the participatory co-design of smart cities requirements and use cases with a focus on end-user requirements, considering applications such as environmental monitoring, water management, goods and people transportation and mobility, waste and public space management.
- To develop a set of IoT enablers that are tailored to the specific circumstances at African cities and integrate them following the reference model to build up a cost-effective smart city platform.
- To leverage the integrated smart city platform and co-created use-cases through the development of web and mobile applications that should be used for validation purposes.
- To validate the deployed smart city platform in real-world scenarios involving end-users and assessing their experience.
- To foster communities of engaged citizens and stakeholders that become part of the co-creation activities for future smart city making.
- To disseminate the results and lessons learnt mainly among sub-Saharan countries stakeholders (policy makers, innovation hubs, entrepreneurs, etc.) but also through scientific and standardization for mainly contributing on the identification and definition of requirements specific for cities in developing countries that shall influence solutions aiming at global applicability.
- To develop business models for the smart city platform and its services that fosters its sustainable adoption and exploitation during and beyond the project.