Game-Theoretic, Secure Resource Allocation for Integrated Satellite-Terrestrial Networks, BWAC Core Project
This project investigates efficient, secure access control in the emerging Cloud-based Integrated Satellite and Terrestrial Networks (CTSN) technology. Game-theoretic approaches are used to solve the complex optimization problem of access control, where baseband processing, resource allocation and interference management are centralized at the cloud. The proposed Stackelberg game between the attacker and the defender of the CTSN provides resiliency to the DoS or hardware attacks to VMs in the cloud. The resulting solution or equilibrium of the proposed game will provide an efficient and fair access control policy, protecting the system from disruption of targeted VMs, while maintaining Quality of Service (QoS), energy, bandwidth, and cost efficiency.
Sponsor
Broadband Wireless Access and Applications Center (BWAC) - Research Site at NCSU
The grant—running from October 1, 2022 to September 30, 2023—is for a total of $50,000.
Principle Investigators
Alexandra Duel-Hallen
More Details
While terrestrial networks are ubiquitous in urban, densely populated areas, sparsely populated regions often suffer from insufficient wireless coverage. To alleviate these service gaps, a series of satellite constellation projects, such as SpaceX, OneWeb, and O3b, are being developed. Moreover, the integration of satellite and terrestrial networks (ISTN), including ‘multi-transport’ systems, which combine GEO/LEO/Terrestrial wireless capability in a single terminal, have recently attracted attention of researchers and the standards organizations. These integrated systems can improve reachability and service continuity in unserved/underserved areas, enhance reliability through connectivity between various access technologies, and increase network resilience and dependability in responding to natural and man-made disasters.
Cloud-based ISTN (CTSN), where baseband processing, resource allocation, and interference management are centralized at the cloud, provide efficient ISTN implementation. A number of virtual machines (VMs) model limited computational capacity of the cloud resource pool. Access control, i.e., assignment of each user to satellite or terrestrial service and to specific VMs, is a complex optimization problem. We propose to investigate game-theoretic approaches to solving this problem. Both cooperative and non-cooperative games will be considered. The player set will contain the cloud operator and the CSTN users, and the players’ objectives and constraints will be determined by the data rates, the Quality of Service (QOS) (delays), energies, and prices. The resulting solution or equilibrium of the proposed game will provide an efficient and fair access control policy.
However, the above access control assumes an ideal, secure ISTN while in CSTN the network functions continuously interact, creating vulnerable scenarios. For example, traffic congestion due to a denial of service attack (DoS) can rapidly deteriorate the control plane services, thus threatening the network availability and causing disruption of targeted VMs. Furthermore, a distributed DoS (DDoS) attack can be stealthily launched from many sources simultaneously. Moreover, malicious destruction of VM hardware (e.g., circuit boards, memory units, and communication ports) can cause severe and lasting damage to the control plane services. Under these attacks, the ideal resource allocation might easily fall short of the targeted performance goals. We propose to develop access control strategies for CTSN that provide resiliency to the DDoS or hardware attacks to VMs in the cloud.
In the proposed Stackelberg game between the attacker and the defender of the CSTN, the actions of the players are modeled as discrete investment levels, which indicate the levels of effort and the resulting chances of success of attack and protection of each VM. The attacker aims to disrupt the system operation, e.g., reducing the data rates and increasing the delays, the energies, and the prices while the defender aims to maintain these metrics as close to the ideal values as possible. The defender invests in tamper-resistant devices, intrusion monitoring, threat management systems that combine firewalls and anti-spam techniques, devices or software that ensure authorized and authenticated access via increased surveillance, scale-up, etc. Both players are under cost constraints. The game equilibrium will provide the defender with the resource allocation strategy that maintains the QoS, energy, bandwidth, and cost efficiency under attacks.
In summary, this project investigates efficient, fair, and secure access control for emerging CSTN technology.