Talar Atechian

Privacy using mobile cloud computing

Crowdsourcing mobile applications are of increasing importance due to their suitability in providing personalized and better matching replies. Crowdsourcing location based services inherit the querying mechanism from its legacy predecessors and this is where the threat lays. In this paper we are going to prove mathematically that the mechanism with highest privacy level “Footprint” does not satisfy the privacy level needed by crowdsourced location based services and then prove that higher privacy level and lower overhead is achieved when these services are implemented in Operator centric mobile cloud architecture.

A new software-based service provision approach for vehicular cloud

Vehicles belonging to a Vehicular Ad hoc NETwork (VANET) can form an ad hoc vehicular cloud in order to provide and to share services between the vehicles. These services may be divided into two categories: (1) Infrastructure as a Service (IaaS) describing physical resources shared as services such as data storage, sensor data, network and computing services; and (2) Software as a Service (SaaS) describing applications that can be run as services in the network. In this paper, a SaaS is an infotainment application requiring multiple infrastructural services such as sensor data. In this context, we propose a service provision mechanism well adapted for VANET. The mechanism aims to select convenient service provider vehicles taking into account the mobility of the vehicles. The mechanism is evaluated by a series of experimentations on OMNeT++ network simulator. The results show a high service delivery ratio representing a satisfactory result for the simulated case study.

Synchronized region based clustering for energy saving in Wireless Sensor Networks

Increasing the life time expectancy is one of the major chalenges in Wireless Sensor Networks (WSN)s. Energy consumption in WSNs is mainly due to the sensing and the transmission capabilities of the sensor nodes. In this paper, we propose an energy-saving multi-hop clustering which balances the position of powerful nodes in the sensor network and hence reduces the total energy dissipation of the nodes. Knowing the base station position and having a variable number of sensor nodes that are deployed randomly, the nodes are clustered using distinct synchronized regions, where a Cluster Heads (CH) is being specified in each region. Instead of using a circle to each cluster, we found out that it is more effective to use a rectangle which cover all the area and take all the spaces of the area into consideration, allowing all the nodes to participate in the transmission. We introduce the location based routing protocol to indicate the location of each node which is available by communicating with a satellite using GPS when nodes are equipped with a small low power GPS receiver. Simulation result of our proposed approach shows that the new algorithm can reduce energy consumption comparing to LEACH protocol by up to 20 percent and obviously improves the network lifetime.

Optimizing Radio Resource Allocation in Multicast Mobile Networks

Mobile cloud computing is a rapidly increasing technology where users demand, for cloud contents, might exceed the capability of mobile networks even with 5G. Multicast communication can offer excellent service for multiple applications, but in some scenarios its efficiency can drop to reach that of unicast. In this work, we will propose a mechanism to optimize the allocation of radio resources when used for multicast, thus offering the mobile operator an opportunity to increase profit without affecting user satisfaction.

Enhanced Revenue Optimizing SLA-Based Admission Control for IaaS Cloud Networks

As cloud markets get mature, CSPs (Cloud Service Providers) will strive to maintain their market shares by ensuring high customer satisfaction which is severely affected by each SLA (Service level agreement) breach. A CSP has to compromise between long-term (maintaining quality of service and SLAs) and short-term (decreasing the rejection of incoming jobs) profitability. In this paper we propose an admission control mechanism for cloud networks that increases the long-term profits without affecting the CSPs short-term targets and its profits enhancements ranges between 80% and 400%.

Micro-economy effect on cloud federation

Cloud federation is considered the future of cloud providers, and for its importance, many researchers proposed federation mechanisms and automatic pricing schemes. To the best of our knowledge, all the proposed federation mechanisms fall in the same fatal hazard by selecting very far cloud providers to establish federation with and this severely decrements their customer satisfaction. In this paper, we raise the awareness on the hazards of price-based cloud selection mechanisms and propose views on how to solve the discussed hazards.