Informatics Publishing Limited

Tharani Vimal ¹, S. A. Sahaya Arul Mary ²

Affiliations
1 Department of CSE, Manonmaniam Sundaranar University, Thirunelveli - 627012, Tamil Nadu, IN
2 Department of CSE, Jayaram College of Engineering and Technology, Trichy - 621014, Tamil Nadu, IN

Abstract

Background: Caching technique is widely used to cope up with the resource constraints and to improve the data accessibility in MANETs. The key constraints are reducing the data access time, overhead and energy cost. Methods: This work proposes a Popularity-driven Caching and Mobility adaptive query Searching (PCMS) to achieve the constraints. The PCMS consists of Cache Node selection, popularity and data size driven replacement (POP) algorithm and mobility adaptive searching for cache data. With the aim of reducing the access delay of popular data, the PCMS selects a node that is closer to the destination in a route to cache the frequently requested data item and serves the future requests. The PCMS takes into account the data popularity and size to decide caching and replacement policy and thus significantly improves the performance of caching. The mobility adaptive searching for cache data looks for an item in Query Directory (QD) nodes and redirects the request to the concerned caching node. It updates the cache information at QD by adhering to the mobility prediction scheme. Findings: The PCMS considers the data popularity and size as key factors to decide caching and replacement policy for the better utilization of cache memory. The PCMS has successfully reduced the network traffic and improves the Cache Hit Ratio by adhering the mobility prediction based update scheme. Applications/Improvements: The proposed system achieves appreciably better performance in simulation than the existing caching systems by utilizing minimum battery power, bandwidth and access delay.

Keywords

Cache Replacement, Mobility Adaptive, Mobile Ad Hoc Network, Popularity Based.

Full Text

   

R. Shankar ¹, Tharani Vimal ²

Affiliations
1 Department of Computer Science and Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, Andhra Pradesh, IN
2 Department of Information Technology, St. Peter’s College of Engineering and Technology, Avadi, Chennai, Tamil Nadu, IN

Abstract

With the emergence of accessing Smartphones in day-to-day life, Mobile Cloud Computing (MCC) technology has become popular with the advantage of resolving the resource constraints in mobile devices through the offloading method. The existing models have presented the different resource allocation solutions to ensure the seamless execution of the applications for the resource-constrained mobile devices with the Quality of Service (QoS). The optimization of resource allocation is the process of potentially allocating remote resources to mobile users without violating the Service Level Agreements (SLAs). However, resource allocation is still becoming a major constraint in the Mobile Cloud (MC) data centers due to higher consumption of energy and time factors during the execution of mobile requests on the remote cloud. The consumption of the energy and response time of the offloaded tasks or applications heavily relies on the cloud resource allocation for the mobile users. Hence, Resource Allocation Optimization (RAO) emerged as the significant objective to select the appropriate cloud resources for the requested tasks to increase the lifetime of the devices with improved time efficiency. Thus, this work focuses on optimizing MC resource allocation by optimizing the allocation of both the computation and communication resources. The proposed RAO model considers two potential factors, such as the energy and response time while allocating the computational and communicational resources. Initially, the Energy and Response time-driven RAO (EARO) approach prioritizes the request generated from the mobile users based on the estimated execution time. Modeling the Estimated Communication and Execution Time (ECET) algorithm tends to allocate the cloud resources and accomplish the minimal response time of the application requests. The EARO approach intends to minimize the execution time as well as the response time towards the target of alleviating the energy consumption during the resource allocation. Moreover, it selects the resources for the inter-VM communication with the knowledge of the minimal migration time ensuring bandwidth resources. Thus, EARO preserves the device's energy with minimal application completion time. The experimental results illustrate that the time efficiency of the proposed EARO model outperforms the existing resource allocation model in the MC environment.

Keywords

MCC, Resource Allocation, Computation, Communication, Optimization, Energy Consumption, Bandwidth, Response Time.

Full Text