On Jamming and Black Hole Attacks in Heterogeneous Wireless Networks

The proliferation of radio-enabled equipment, including sensors, controllers, actuators, mobile handhelds, etc., into the today's heterogeneous wireless network (HWN) has led to the support of different vertical services on-top of the same physical infrastructure. Such an approach increases the potential for enhanced multiplexing gains due to the joint utilization of multiple network tiers and their resources. However, it also links the availability of new services to the robustness of the HWN infrastructure against malicious attacks. In this paper, we study the impact of two prominent denial of service (DoS) attacks, the jamming and the black hole attack, on network coverage of multi-tier HWNs. In particular, we consider a HWN model of multiple tiers, where the locations of nodes belonging to each tier are modeled by a homogeneous Poisson point process (PPP) with known intensity. Each tier consists of regular nodes, jammers (disrupting downlink communications by transmitting at constant power), and black holes (appearing as regular nodes but without forwarding received packets to their legitimate destination). The type of a node for a given network tier is determined by a prescribed probability, whereas jammer and black hole nodes are considered to act independently from each other. We further focus on the challenging scenario where end users can access a limited number of network tiers, while we also allow different tiers to potentially utilize different spectrum bands. Assuming that successful service reception from a node belonging to a given network tier requires a minimum received signal quality threshold to be attained, we derive exact expressions and performance bounds on the coverage probability of random multi-tier HWNs with joint jamming and black hole attacks, which depend on the capability of network nodes to detect and avoid association with malicious nodes (i.e., jammers or black holes). Detailed numerical results highlight the usefulness of the proposed analysis, providing valuable insights on system design and parameterization towards enhanced network robustness.