A Load-Balanced Re-embedding Scheme for Wireless Network Virtualization
ABSTARCT :
In wireless network virtualization processes, load imbalance in substrate nodes and substrate links may significantly reduce the number of virtual network requests served by a substrate network. In addition, load imbalance also impairs resource utilization and increases embedding cost of the virtual networks. To tackle these issues, this work proposes a re-embedding scheme for virtual node (RSVN). Based on the load condition of a substrate network, RSVN first determines whether to start re-embedding procedure or not, and then selects proper re-embedded virtual nodes according to their re-embedding factors. Next, RSVN re-embeds the selected virtual nodes to balance the loads of a wireless substrate network. The extensive simulation results showed that RSVN improves the acceptance ratio of virtual network requests, increases the resource utilization of substrate networks, and reduces the average embedding cost of the entire virtual networks.
EXISTING SYSTEM :
? Network Virtualization allows the co-existence of multiple Virtual Networks (VNs) on the same physical infrastructure (often referred as substrate network, SN).
? Virtual Network Embedding (VNE) maps the virtual nodes and links requests from users onto the graph representing the physical infrastructure nodes and links and their connectivity.
? The algorithm focuses on virtual nodes requiring more resources by extending their allocations and maintaining their connectivity (even if the node is migrated) to other resources while tidying up (or consolidating) the infrastructure. The algorithm outperforms existing approaches.
? Unfortunately, they lead to network instability and service disruptions.
DISADVANTAGE :
? We identify a new flow scheduling problem in big data centers in clouds, i.e., dynamical load-balanced scheduling (DLBS), and formulate the DLBS problem.
? We propose a set of heuristic scheduling algorithms to address the DLBS problem.
? We firstly formulate the DLBS problem, and then develop a set of efficient heuristic scheduling algorithms for the two typical OpenFlow network models, which balance data flows time slot by time slot.
? Load-balanced flow scheduling for big data centers in clouds, in which a large amount of data needs to be transferred frequently among thousands of interconnected servers, is a key and challenging issue.
PROPOSED SYSTEM :
• To circumvent this problem, some metrics for AP association that define a relation between RSSI and the amount of associated stations to an AP were proposed.
• Alternative approaches have been proposed, which perform a load balancing among APs by including load conditions in the frames in order to allow the STA to select the least loaded AP.
• In the proposed mechanism, the STAs manage associations with multiple APs using one virtual wireless interface for each AP.
• The performance of the proposed load balancing mechanism is closely tied to the active time duration of each virtual interface.
• Therefore, the proposed virtualization mechanism provides a dynamic load balancing across the APs while performing a fair resource sharing between STAs.
ADVANTAGE :
? Load balancing means that all resources in a system are equally shared by all tasks in some measures. It can be mathematically described by means of a performance criterion.
? The performance of mid-way load-balanced scheduling can be measured by the network throughput under realistic conditions in the two network models.
? Thus, we can conclude that our DLBS can efficiently balance the global load so that it significantly improves throughput, transmission delay and bandwidth utilization rate especially under non-uniform network transmission patterns.
? This system can adaptively schedules a multi-stage switching fabric to efficiently utilize aggregated network resources.
|