This paper presents a solution to increase bandwidth utilization and to reduce the time necessary to recover from network failures in multi-domain bandwidth on demand service provisioning. The proposed solution is based on software-defined networking (SDN) in order to implement novel traffic engineering (TE) techniques. In fact, most standards development organizations and researchers agree that the logically centralized control plane and high programmability of SDN can revolutionize TE with novel and powerful strategies. In this context, the dynamic path computation (DynPaC) framework, relies on the OpenFlow SDN protocol to provide resilient layer 2 services with bandwidth guarantees within OpenFlow domains. In order to make DynPaC support multi-domain service provisioning, it has been extended with an additional operational mode and REST API calls, making it fully compliant with the Network Services Interface-Connection Service protocol. In this way, DynPaC can behave as a network resource manager (NRM) in the network services framework. As a result, the presented solution is the first OpenFlow-enabled NRM with resiliency capabilities that implements novel TE strategies that enhance the utilization of network resources. In order to validate the proposal, it has been implemented using the ONOS network operating systems and several tests have been conducted using diverse service demand patterns. The obtained results show that DynPaC is suitable to be used as an NRM and that the number of accepted services increases thanks to the flow reallocation mechanism included in the framework, achieving in this way a better utilization of network resources.

Higher interconnectivity among devices, machines, the cloud and humans is envisioned in the actual trend of automation, also known as Industrial Internet of Things (IIoT). These industrial control systems, which may require high availability and/or safety related capabilities, are no longer isolated from the corporate environment or Internet. Software updates will be needed during the product life cycle, due to the long service life, the increasing number of security related vulnerabilities discovered on these industrial control systems and the high interconnectivity desired in IIoT. These updates aim at fixing all these security weaknesses, bugs and vulnerabilities that could appear, while the required safety integrity levels are ensured. Security-related concerns have just been addressed by the safety engineering community, because of the increasing number of cyber-attacks against safety-critical systems, such as Stuxnet. Moreover, system shut-downs caused by software updates could not be plausible when high availability is required. Typically, in order to perform the software update, the whole industrial process or the production is halted, so that the software upgrade is safely applied. However, this scenario might not be applied in critical infrastructures, such as nuclear or hydro- electrical power plants, where these production and service interruptions are not acceptable from the business and service point of view. This article presents an analysis of existing dynamic software updating techniques, which may be applied for safe and secure industrial control systems. These techniques aim at updating the running code, without the need of a halt and restart, increasing the availability of the industrial system.

Security is a key aspect in the development of innovative and valuable services based on Cyber-Physical Systems (CPSs). In the last years, the research area related to CPS security has received a significant attention, dealing with the design of different architectures, security protocols, and policy models. However, beyond monitoring data publishing behavior, CPSs are expected to offer some manageability-related services, and the proper fine-grained and flexible access control model remains challenging due to both criticality and feasibility. In fact, traditional security countermeasures cannot be applied directly to any sensor in CPS scenarios, because they are too resource-consuming and not optimized for resource-deprived devices. Different access control models facing both feasibility and enforcement tightness are arising as a way to solve the mentioned issues related to resource limitations, and this study provides a deep survey on them.

Software Defined Networking (SDN) is a new network paradigm that decouples the control from the data plane in order to provide a more structured approach to develop applications and services. In traditional networks the routing of flows is defined by masks and tends to be rather static. With SDN, the granularity of routing decisions can be downscaled to single TCP sessions, and can be performed dynamically within a single data stream. In this context We propose a novel approach – coined as micro flow aware routing – aimed at implementing routing of flows based on the properties of transport-level information, which is closely related to the type of application. Our proposed scheme relies on the early characterization of the flow based on statistical predictors, which are computed over a time window spanning the first exchanged packets over the session. We evaluate different window lengths over real traffic data to examine the Pareto trade-off between the earliness of flow characterization and its predictive accuracy. These results stimulate further research towards ensuring the practicality of the scheme.

Network monitoring is a paramount aspect for the detection of abnormal and malicious activity. However, this feature must go hand by hand with mitigation techniques. On SDN environments, control techniques may be easily developed as a result of its ability for programming the network. In this work, we take advantage of this fact to improve the network security using the sFlow monitoring tool along with the SDN controller. We present an architecture where sFlow is in charge of detecting network anomalies defined by user rules, while the SDN technology is responsible to mitigate the intrusion. Our testbed has been implemented on Mininet and the SDN environment is governed by Opendaylight controller and the OpenFlow southbound protocol. Experimental validation demonstrate that our system can effectively report various types of intrusion associated with the reconnaissance phase of an attack.