No abstract

No abstract

No abstract

Fog computing has been prioritized over cloud computing in terms of latency-sensitive Internet of Things (IoT) based services. We consider a limited resource-based fog system where real-time tasks with heterogeneous resource configurations are required to allocate within the execution deadline. Two modules are designed to handle the real-time continuous streaming tasks. The first module is task classification and buffering (TCB), which classifies the task heterogeneity using dynamic fuzzy c-means clustering and buffers into parallel virtual queues according to enhanced least laxity time. The second module is task offloading and optimal resource allocation (TOORA), which decides to offload the task either to cloud or fog and also optimally assigns the resources of fog nodes using the whale optimization algorithm, which provides high throughput. The simulation results of our proposed algorithm, called whale optimized resource allocation (WORA), is compared with results of other models, such as shortest job first (SJF), multi-objective monotone increasing sorting-based (MOMIS) algorithm, and Fuzzy Logic based Real-time Task Scheduling (FLRTS) algorithm. When 100 to 700 tasks are executed in 15 fog nodes, the results show that the WORA algorithm saves 10.3% of the average cost of MOMIS and 21.9% of the average cost of FLRTS. When comparing the energy consumption, WORA consumes 18.5% less than MOMIS and 30.8% less than FLRTS. The WORA also performed 6.4% better than MOMIS and 12.9% better than FLRTS in terms of makespan and 2.6% better than MOMIS and 4.3% better than FLRTS in terms of successful completion of tasks.

Banana (Musa spp.) is one of the most valuable global agricultural commodities, with commercial plantations responsible for supplying nearly 15% of total global banana production. These plantations are underpinned by major infrastructural investments and a high dependence on fertilizer, pesticide and irrigation inputs. In contrast, smallholders and subsistence farmers often cultivate bananas for local markets with minimal inputs. Water stress due to increasing rainfall variability and competition for water resources are emerging as major production constraints for both commercial and smallholder production. Water stress-induced yield losses of up to 65% have been reported due to loss in bunch weight even in moderate to low rainfall areas. Thus, investments in more efficient irrigation systems and water-saving technologies are being widely promoted to increase water productivity through improved scheduling to reduce drainage and runoff losses. This paper synthesises scientific and industry evidence on crop growth and development including root and shoot development, plant water relations, and yield response to water. It also critiques the importance of irrigation scheduling for maximising irrigation efficiency. New evidence to support the synchronization of irrigation with crop water demand to reduce environmental impacts is provided. High variability in crop water demand (1200 to 2690 mm per year) was

No abstract

No abstract

Wireless sensor networks (WSNs) have received much attention in recent years because of its broad area of applications. In the same breadth, it also faces many challenges. Time synchronisation is one of those fundamental and important challenges faced by WSN. Several approaches have been proposed in recent years for time synchronisation. Most of the approaches are based on synchronising to a reference (root) node's time by considering a hierarchical backbone for the network. However, this approach seems to be not purely distributed, higher accumulated synchronisation error for farthest node from the root and subjected to the root node failure problem. In this study, a fully distributed time synchronisation algorithm is proposed which synchronises all nodes' time to a consensus value, that is, an average of all nodes' initial clock values. The algorithm exploits a selective pairwise averaging approach with a linear message complexity of O(n), to achieve faster convergence and better synchronisation accuracy than the existing random pairwise average time synchronisation protocol. Simulation results show that almost 50% improvement is achieved in convergence speed, measured in terms of number of iterations, and nearly 30% improvement in total synchronisation error.