Software Defined Networks for the Electric Power and Energy Systems

In recent years, Software Defined Vehicular Networks (SDVN) [1], have been emerged as a promising technology to simplify network management and enable innovation through network programmability. Thus, it gains significant attention from academia and industry. The SDN technology allows the decoupling of control and data planes which provides: (i) an abstraction for vehicular applications to the underlying networking infrastructure, and (ii) a logically centralized networking intelligence and network state.

The convergence of SDN with Internet of things or mobile devices such as vehicles is seen as an important direction that can address most of the mobile networks current challenges [2]. The use of SDN’s prominent features provides the needed support to vehicular network applications to enhance the user experience. Moreover, the SDN features can meet the advanced demands of vehicular ad hoc network  like high throughput, high mobility, low communication latency, heterogeneity, and scalability. In particular, the inherent features of SDN include: (i) flexibility through dynamic programmability on networking elements, (ii) support for heterogeneous applications through network virtualization, and (iii) efficient management of various services using centralized global network knowledge. These features of SDN could help to ensure secure and efficient deployment of different services.

In a smart grid application, the SDVN architecture can be used where data plane includes Electric Vehicles (EVs) and Electrical Vehicle Supply Equipments (EVSEs). Attacks on the smart grid include network flooding, topology poisoning, and transmission jamming. When an electric vehicle needs to connect to the EVSEs, an information message is sent to the controller in order to track the current network topology and status of the network. Moreover, anomaly detection systems such as the ones in [3] can be added to the controller to monitor network traffic and detect compromised data. The controller can install forwarding rules, and it can detect attacks implied by unusual behaviors in the smart grid application.

Although various architecture designs for SDVNs have been proposed in the literature to improve the communication reliability and security in  mobile electric vehicles scenarios, the comprehensive investigation to evaluate the deployment feasibility, effectiveness, and correctness of these architectures remains an open issue. In particular, the new security and privacy vulnerabilities that arise due to the coupling of new technologies (such as SDN, NFV, and mobile edge computing) with the existing smart grid applications should be carefully studied, and the same is completely missing so far. For instance, researchers should not only report the benefits of using SDN to improve VANET architecture, but the new issues (e.g., service latency, mobility, and securing the SDN controller) that are inherent to SDN and now hindering the performance of the SDVNs should also be investigated and discussed.

 

---

References:

[1] W. Ben Jaballah, M. Conti, C. Lal, “Security and Design Requirements for Software-Defined VANETs”, Elsevier Computer Networks, vol. 169, March 2020.

[2] A. Ydenberg, N. Heir, and B. Gill, “Security, sdn, and vanet technology of driver-less cars,” in 2018 IEEE 8th Annual Computing and Communication Workshop and Conference (CCWC), Jan 2018, pp. 313–316

[3] S. Shin, P. Porras, V. Yegneswaran, M. Fong, G. Gu, and M. Tyson “Fresco: Modular composable security services for software-defined  networks,” in IEEE Network and Distributed System Security Symposium (NDSS), 2013.