Ssu-Chi Kuai

An Energy-Efficient Sensor Deployment Scheme for Wireless Sensor Networks Using Ant Colony Optimization Algorithm

Sensor deployment is one of the most important issues in wireless sensor networks (WSNs), because an efficient deployment scheme can reduce the cost and enhance the detection capability of the WSNs. Due to packet forwarding, sensors closer to the sink consume more energy than those farther away. In this paper, we propose a sensor deployment scheme, which can achieve full coverage of the monitoring area and prolong network lifetime. We consider a real world situation where the initial energy of the sensors is different from each other. First, to achieve full coverage using as few sensors as possible, we compute the average angle between the sensor nodes. Then, we provide two methods to achieve energy balance. In the first method, we propose a sweep-based scheme to move the sensors as requested. In the second method, we transform the deployment problem into the multiple knapsack problem and based on ant colony optimization algorithm, we propose a deployment strategy to improve the network lifetime.

A Receiver-Initiated MAC Protocol with Packet Train Design for Underwater Acoustic Sensor Networks

An efficient MAC protocol can reduce the collision and increase the throughput of the wireless sensor networks. Due to the long propagation delay and the low data rate, many transmissions of control packets will occupy a lot of channel resources, power and transmission time. In this paper, we propose a receiver-initiated MAC protocol with packet train for UWSN called multi-receiver MAC protocol (MR-MAC). Our protocol will increase of the throughput and reduce of the transmission of control packet for handshaking. Our MR-MAC protocol can make more than two nodes to communicate in one handshake held by a main receiver. By scheduling the packet transmission time, the data packet will be sent in a packet train manner and the receiver can receive data packet without collision. Our simulation results show that our protocol can increase the throughput and decrease the packet delay than the other existing MAC protocols.

A receiver-initiated MAC protocol for underwater acoustic sensor networks

Due to the long propagation delay and the low data transmission rate, many transmissions of control packets will occupy a lot of channel resources, power consumption and transmission time for underwater acoustic sensor networks (UWSNs). In this paper, we propose a receiver-initiated MAC protocol with packet train for UWSNs called multi-receiver MAC (MR-MAC) protocol. Our protocol will increase of the throughput and reduce of the transmission of control packet for handshaking. Our MR-MAC protocol can make more than two nodes to communicate in one handshake held by a main receiver. By scheduling the packet transmission time, the data packet will be sent in a packet train manner and the receiver can receive data packet without collision. Our simulation results show that our protocol can increase the throughput and decrease the packet delay than the other existing MAC protocols.

Auto-scaling Strategy for Amazon Web Services in Cloud Computing

Auto scaling mechanisms have become a typical paradigm in cloud computing environments. Such mechanisms can increase or minimize the number of virtual machines according to user demands, consequently achieving pay-per-use objectives. However, auto scaling mechanisms provided by infrastructure-as-a-service providers must strictly follow user-defined thresholds, the drawback of such mechanisms is that they cannot respond to real-time Internet traffic loads by following user-defined thresholds. Therefore, we propose a dynamic threshold adjustment strategy that can expedite the creation of virtual machines according to workload demands. The proposed strategy can reduce the web application response time and error rate when the system is under a heavy workload. In addition, it can expedite the release of virtual machines to reduce virtual machine running time when the system is under a light workload. According to our experimental results, we found that CPU-intensive web applications require an excellent threshold control strategy. Therefore, the proposed strategy can satisfy this requirement by effectively reducing the response time of applications, virtual machine running time, and error rate.

Dynamic Load-Balancing Mechanism for Software-Defined Networking

Software-defined networking (SDN) is a new network model, and it changes the current limit network facilities. In SDN, the control logic is separated from network devices such as routers and switches. Accordingly, network devices only need to transfer data (data layer), while the control logic is handled by a central controller. In this paper, we propose a dynamic load-balancing mechanism to reduce response time by adjusting rules in the SDN flow table. Our scheme can adjust the binary tree depending on the load on each server. When a server is out of capacity, it can dynamically transfer excess traffic load to servers with less load. Simulations show that our scheme not only balances server load but also maximizes server usage.

A Resource Provision Strategy for Software-as-a-Service in Cloud Computing

Abstract The business model of Software-as-a-Service (SaaS) is gradually becoming more refined. Reducing the operation cost of cloud service providers while maintaining service efficiency has become an important topic. The present study focuses on suggesting a resource provisioning strategy for load changes during the operations of an SaaS cloud service provider (CSP). The strategy is to suggest corresponding provision portfolio configurations that can reduce operation cost and maintain efficiency of applications based on surplus resources after reducing the use of virtual machines rented for the SaaS CSP. The underlying concept is identical to memory paging management in operating systems: minimizing fragmented space to utilize resources effectively. Testing results confirm that the strategy suggested in this study can effectively configure a provision portfolio with reduced operation cost. Compared with methods suggested by IaaS CSPs in the mainstream market, the method proposed in the present study can reduce the hourly rental of virtual machines by 40% and enable SaaS CSPs to realize considerable profit by reducing cost.

Efficient scheduling algorithm for multiple mobile subscriber stations with unsolicited grant service in an IEEE 802.16e network

The IEEE 802.16e standard introduces the concept of mobile subscriber stations (MSSs) to provide mobility support. Five quality of service (QoS) classes have been defined to meet the QoS requirement of different connections between a base station and a subscriber station. Among these QoS classes, unsolicited grant service (UGS) has been designed to support real-time service flows that periodically generate fixed-size data packets. Most of the existing scheduling schemes for a UGS class consider scheduling only a single MSS; even if such schemes consider multiple MSSs, the QoS requirement of such MSSs is not appropriately satisfied after scheduling. This paper proposes a scheduling scheme, which schedules multiple MSSs with UGS connections so that the QoS requirement of each MSS can be satisfied after scheduling. The simulation results showed that the proposed approach achieved a bandwidth utilization of more than 90% and did not incur a high bandwidth waste when the number of connections was high.

An efficient load balance data collection scheme in wireless sensor networks

Wireless sensor networks (WSNs) are used in a wide range of applications. In WSNs, sensor nodes transmit the data to sink via one-hop or multi-hop. The sensor nodes near the static sink are overloaded than others. By mobile sink, we can improve the network performance and then balance the energy consumption. It is important to use the mobile sink balancing the loading among sensor nodes and helping the adjusting of network lifetime. A set of sensor nodes are selected as rendezvous points (RPs), and non-RP nodes forward their data to RPs. In this paper, we present an efficient load balance data collection scheme to choose the RPs plain the path of mobile sink to visit all this RPs. The LBDC algorithm prolongs the lifetime of sensor network and balances each RPs amount of children. Our simulation results show that our scheme can increase the network lifetime.

A QoS data collection scheme for wireless sensor networks

In wireless sensor networks (WSNs) the use of mobile sink has been recently attracting more attention. After the network working for some time, the sensor data which is generated must be dropped because of the sensor nodes of storage is limited. The dropped of sensing data will affect the completeness of quality of information (QoI). If the network environment uses multi-hop to send sensing data to the sink, the effect of delay will be lower, but the sensor nodes will spend more energy. However, the extant methods are mostly not taking the data drop caused by the generation rate of sensing data into account. Also, some sensing data will violate the delay restrictions and loss its value because of the delay time caused by mobile sink moving in the environment. In this paper, we propose a QoS data collection scheme by using mobile sink to energy-efficiently collect all the sensing data of each sensor node. The tour planning of mobile sink collecting sensor data is planned by the storage of sensor nodes and the latency of application. Our scheme can minimize the distance of the total tour and the total energy cost, as well as to satisfy the energy constraint of the mobile sink. Simulation results show that our scheme can effectively collect sensor data.