In Cloud-based computing, job scheduling and load balancing are vital to ensure on-demand dynamic resource provisioning. However, reducing the scheduling parameters may affect datacenter performance due to the fluctuating on-demand requests. To deal with the aforementioned challenges, this research proposes a job scheduling algorithm, which is an improved version of a swarm intelligence algorithm. Two approaches, namely linear weight JAYA (LWJAYA) and chaotic JAYA (CJAYA), are implemented to improve the convergence speed for optimal results. Besides, a load-balancing technique is incorporated in line with job scheduling. Dynamically independent and non-pre-emptive jobs were considered for the simulations, which were simulated on two disparate test cases with homogeneous and heterogeneous VMs. The efficiency of the proposed technique was validated against a synthetic and real-world dataset from NASA, and evaluated against several top-of-the-line intelligent optimization techniques, based on the Holm’s test and Friedman test. Findings of the experiment show that the suggested approach performs better than the alternative approaches.

Connected cars are equipped with a rich set of sensors, such as GPS, accelerometer, video cameras, and pollution detectors. The information generated by these sensors can be used to offer a wide range of on-demand services, such as congestion notification, parking lots, and video surveillance. These services need a reliable and low latency unicast communication scheme in order to efficiently deliver the information requested by drivers. In this paper, a local traffic aware unicast routing scheme is proposed. To overcome network fragmentation, the proposed scheme relies on base stations and virtual base stations to transmit information from source car to the destination car using backhaul link. In this model, as base stations are sparsely deployed in the junction areas, a car moving in the junction area acts as a virtual base station node to support the routing process in the absence of a base station. Moreover, it avoids the impact of unreliable channel on information delivery. In the proposed scheme, each base station and virtual base station uses the short status messages (beacons) exchanged by the cars to form a local database of car locations. The stored information is used to find a base station or virtual base station that offers a minimum delay path to the destination car. The simulation results show that the proposed scheme outperforms the existing routing schemes in terms of end-to-end delay and packet delivery ratio. The proposed scheme is also validated by a connected car prototype built in an indoor laboratory environment.

In recent years, Industrial Wireless Sensor Network (IWSN) is gaining more popularity due to many applications in industries like fire detection, hazardous gas leakage detection, temperature monitoring, localization of sensors, etc. However, faulty sensors in the network may degrade the performance of the applications. In this paper, a software defined network (SDN) based fault detection method is proposed for IWSN. In this method, SDN plays an important role for controlling the whole system by setting a fault detection algorithm at the cluster heads (CHs). The CH periodically receives the monitoring data from the sensors and follows the fault detection algorithm set by the SDN to detect the faulty sensors in the network. The fault detection algorithm uses a statistical trimean method to detect the faulty sensors. Simulation results show that our proposed method performs better than Ji's fault detection method in terms of detection accuracy (DA) and false alarm rate (FAR). A IWSN prototype is also designed to evaluate the performance of the proposed method.