Mustafa Ozger

Event-driven spectrum-aware clustering in cognitive radio sensor networks

Wireless sensor networks (WSN) with dynamic spectrum access (DSA) capability, namely cognitive radio sensor networks (CRSN), is a promising solution for spectrum scarcity problem. Despite improvement in spectrum utilization by DSA capability, energy-efficient solutions for CRSN are required due to resource-constrained nature of CRSN inherited from WSN. Clustering is an efficient way to decrease energy consumption. Existing clustering approaches for WSN are not applicable in CRSN and existing solutions for cognitive radio networks are not suitable for sensor networks. In this paper, we propose an event-driven clustering protocol which forms temporal cluster for each event in CRSN. Upon detection of an event, we determine eligible nodes for clustering according to local position of nodes between event and sink. Cluster-heads are selected among eligible nodes according to node degree, available channels and distance to the sink in their neighborhood. They select one-hop members for maximizing the number of two-hop neighbors that are accessible by one-hop neighbors through cluster channels to increase connectivity between clusters. Clusters are between event and sink and are no longer available after the end of the event. This avoids energy consumption due to unnecessary cluster formation and maintenance overheads. Performance evaluation reveals that our solution is energy-efficient with a delay due to spontaneous cluster formation.

Event-to-Sink Spectrum-Aware Clustering in Mobile Cognitive Radio Sensor Networks

Cognitive radio sensor networks (CRSNs) are event-based systems such that sensor nodes detect events and the event readings of the sensors are collaboratively conveyed in a multi-hop manner through vacant channels from event regions to a sink. Hence, the event-to-sink communication and the dynamic radio environment require a coordination scheme in CRSNs. In this paper, we propose a spectrum-aware clustering protocol to address the event-to-sink communication coordination issue in mobile CRSNs. Our clustering scheme consists of two phases. The first phase is the determination of nodes eligible for clustering, and the second phase is to form clusters among those nodes according to vacant spectrum bands. Clusters are temporary and they are not preserved after the end of events. Furthermore, we find average re-clustering probability, expected cluster coverage area, and find maximum event generation frequency for energy-efficient operation of our protocol. We study performance of our protocol in terms of control and data packet exchange, time steps required for clustering, connectivity of clusters, energy consumed for clustering, and re-clustering ratio due to the mobility. Performance comparison simulations show that our algorithm has better performance in terms of connectivity and energy consumption.

On the Maximum Coverage Area of Wireless Networked Control Systems With Maximum Cost-Efficiency Under Convergence Constraint

The integration of wireless communication and control systems revealed wireless networked control systems (WNCSs). One fundamental problem in WNCSs is to have a wide coverage area. For the first time in the literature, we address this problem and we obtain the maximum coverage area by solving an optimization problem. In this technical note, we consider a WNCS where the output sensor measurements are transmitted over separate heterogeneous multi-hop wireless ad-hoc subnetworks. The observation process is divided into N parts and the system state is estimated using the Kalman filter. We present the critical arrival probability for a sensor measurement packet such that if the packet arrival probability is larger than the critical value, it is guaranteed that the estimator of the WNCS converges. We derive the maximum total coverage area of the heterogeneous wireless subnetworks having maximum cost-efficiency under the constraint of the convergence of the WNCS estimator.

Energy Harvesting Cognitive Radio Networking for IoT-enabled Smart Grid

The Internet of Things (IoT) provides connectivity to the objects that monitor and sense the environment to integrate physical world with digital world. If IoT is enabled in the Smart Grid (SG), it can benefit from advantages of the IoT such as interoperability, connectivity, etc. By combining the IoT with energy harvesting (EH) and cognitive radio (CR) techniques, the problems of SG, such as harsh channel conditions and limited battery power, may be resolved. Hence, incorporation of EH and CR reveals a new networking paradigm for IoT-enabled SG. To this end, we first introduce CR usage in the IoT-enabled SG, and explain the advantages and challenges of CRs. Furthermore, we propose EH approaches for the resource constraint of wireless devices in the IoT-enabled SG. Operation and node architecture of energy harvesting cognitive radios (EH-CR), and network architecture of the IoT-enabled SG are described to explain details of our networking paradigm. Open issues and future research directions are discussed to enable this new paradigm.

On the maximum coverage area of wireless networked control systems under stability and cost-efficiency constraints

The integration of wireless communication and control systems revealed wireless networked control systems (WNCSs). One fundamental problem in WNCSs is to have a wide coverage area. For the first time in the literature, we address this problem and we obtain the maximum coverage area by solving an optimization problem. In this paper, we consider a WNCS where the output sensor measurements are transmitted over separate multi-hop wireless ad-hoc subnetworks. The system state is estimated using the Kalman filter. We present the critical arrival probability for a sensor measurement packet such that if the packet arrival probability is larger than the critical value, it is guaranteed that the expected state estimation error covariance is bounded, and hence the WNCS is stable. We find the optimum hop-diameter of a multi-hop wireless ad-hoc subnetwork under the constraints of both the stability of the WNCS and the cost-efficiency of the multi-hop wireless network. Furthermore, under these constraints, we derive the maximum total coverage area of the wireless subnetworks. The numerical analyses show that the maximum total coverage area can be increased by appropriately adjusting the number of sensors, the successful packet transmission probability between relay nodes, and the eigenvalues of the system matrix.

Crowdsourcing-based mobile network tomography for xG wireless systems

Network size and number of mobile users are ever-increasing with the advancements in cellular network technologies. Hence, this situation makes the network monitoring highly complex. Although there are numerous network tomography approaches, service providers need real-time network monitoring tools to provide better network utilization. In this paper, we propose a crowdsourcing-based real-time network tomography framework. In the proposed framework, channel condition and user data usage are monitored via an application at the mobile terminals, and then the mobile terminals transmit their data to the server. In this way, the network and user behavior can be continuously monitored, and real-time actions can be implemented to improve the network performance. By using the proposed framework, we propose an optimization framework for the amount and reporting frequency of the transmitted data to avoid battery drain at the mobile terminal and network congestion. At the end, we provide simulation results for the proposed optimization framework.

Maximization of energy-efficiency under convergence constraint in wireless networked control systems

Wireless networked control system (WNCS) is a control system that a wireless network closes the control loop. WNCS estimator, i.e., Kalman filter, estimates the system state according to the observations of sensors. These observations which are from N independent subnetworks are conveyed to the Kalman filter through vacant bands opportunistically with cognitive radio capability of the nodes. We characterize the successful packet delivery probability and study the maximization of energy-efficiency of overall system under the convergence constraint of the Kalman filter by defining an optimization problem. We also find a lower bound on maximum total coverage area. Furthermore, we perform numerical analysis to observe the effects of system parameters such as number of subnetworks, average ON probability of primary users, transmission ranges and densities of sensor nodes and primary users, and false alarm probability.