Department Publications

This paper introduces a novel traffic reduction method for Networked Control Systems (NCS) called Deadband Error Modulation (DEM). The deadband modifies the periodic dynamics of Delta Modulation (DM) based techniques such as our previously proposed Event Triggered Adaptive Differential Modulation (ETADM). Like ETADM it can be used for encoding individual scalar elements of a vector (i.e., the state vector or control input). It is also robust to bounded packet drops.

Most of the intrusion detection and prevention systems (NIDS/NIPS) are utilizing signature based matching which makes use of a stored rule base for distinguishing malicious network traffic. Dynamic rule-base updates has become a requirement due to the increasingly large number of new attack types identified every day, while the downtime associated with rule base updates has become a major drawback in hardware based Network Intrusion Prevention Systems. In addition, the NIPS are facing the challenge of expanding their throughput to cope with the ever growing high bandwidth requirement. This paper presents a novel scalable architecture for a NIPS with stateful packet inspection. The implementation is capable of processing Snort signatures and also analyzing packet streams to detect attacks with TCP segment manipulation. The system ensures run-time rule reconfigurability, which eliminates the risk of …

There exists some common bottlenecks among the many available Software Defined Networking (SDN) controllers. Of these, the current analysis uses the Floodlight software controller to identify bottlenecks. The analysis has implemented these bottlenecks in a Field Programmable Gate Array (FPGA) in order to assess its feasibility. Thereafter, a Hybrid SDN controller has been developed. Generally Hybrid SDN refers to a networking approach where traditional networking and SDN protocols are used and operated in the same environment. Hybrid in this paper context refers to a controller which is developed both on software and hardware platforms in order to achieve a higher scalability which is a current issue existing in software controller, and thereby overcoming the identified bottlenecks. The analysis has offloaded computational intensive models to an FPGA and tested for performance. These performance figures have then been compared against the performance figures as of a stand-alone software controller.

Power control schemes are proposed for a full-duplex (FD) device-to-device (D2D) network operating as an underlay of a primary FD network. As networks with different quality of service (QoS) requirements may coexist in future heterogeneous networks, it is assumed that the underlay D2D network is operating with best effort QoS provisioning, while the primary FD network requires guaranteed QoS. A centralized power control (CPC) solution is proposed by casting the power control problem as a convex optimization problem. Furthermore, a heuristic distributed power control (DPC) solution is proposed to reduce information feedback overhead. It is found that the heuristic DPC achieves D2D user data rate that is within 87% and 80% of that provided by the CPC scheme in 3-node and 2-node FD primary network, respectively.