Task Offloading using Sparrow Search Algorithm

الموضوعات

علوم تطبيقية وتكنولوجية

الملخص

With the exponential growth of the Internet of Things (IoT), computationally intensiveand delay-sensitive applications are rapidly growing. The data generated by these appli-cations such as real-time image processing, augmented reality, video games, and location-based applications need significant computation power. However, IoT devices are lim-ited in terms of battery life, processing power, and storage capacity. Cloud computing issuggested to be utilized in offloading computation-heavy tasks from IoT devices to cloudservers. Although task offloading from IoT devices to the cloud may overcome the limita-tions of IoT devices, Cloud computing might result in high latency which directly affectsthe deliverable and performance of the IoT applications. Fog computing has been devel-oped to address this issue by placing storage and processing components at the edge ofthe network closer to IoT devices. Fog computing can deliver computing resources to IoTdevices with lower latency compared to cloud computing. However, the optimal fog noderesource allocation while minimizing the delay is challenging and must be done efficiently.Therefore, task offloading to fog nodes encountered the challenge of minimizing thee delayfor IoT applications.In some IoT areas such as the Internet of Vehicles (IoV), the challenge becomes moreserious. IoV is one of the most important implementations of the IoT which attempts toensure the safety of the vehicle and its environment in addition to achieving automaticdriving. Applications in the IoV are delay-sensitive and have more strict QoS than thosein other IoT applications. Furthermore, it is more challenging to apply the fog computingarchitecture to IoV networks due to the high mobility of vehicles. On the other hand,the coverage of fog nodes in fog computing is limited geographically and vehicles mightenter areas with poor fog computing coverage. Hence, areas with poor fog computingcoverage can be handled by vehicle-to-vehicle (V2V) task offloading which is challengingto optimize as well. In this thesis, we address task offloading in two different scenarios. The first scenariois the task offloading model for the IoT. This model is designed as a multi-objective opti-mization problem to find the best offloading decision. The objective is to reduce responsetime and total task offloading cost, which include latency cost processing cost, and mem-ory cost. An IoT-Fog architecture has been used to offload the task in this model. Weprovide extensive experiments and comparisons with the standard SSA, ACO, PSO, ABC,and RR algorithms to show the efficacy of the model. The second task offloading sce-nario is a multi-hop task offloading model for the IoV in areas with poor fog computingcoverage. A candidate vehicles selection mechanism is used in this model to select the can-didate vehicles from the neighboring vehicles and ensure that the vehicle can successfullytransmit the tasks to the candidate vehicle for processing. We provide extensive experi-ments and comparisons with the standard SSA, BAT, greedy, GPUPSO, TLBO, ACO, GA,and RO algorithms to show the efficacy of the model. We proposed a Multi-ObjectiveTask Offloading model based on Modified SSA (MOTO-MSSA) for IoT and a Multi-HopTask Offloading based on Modified SSA (MHTO-MSSA) for IoV to address those twochallenges using the modified SSA (MSSA). The results reveal that MOTO-MSSA boostsconvergence speed and exceeds existing techniques in terms of cost and response time withminimum overhead. The results demonstrate that MHTO-MSSA achieves faster conver-gence and outperforms BAT and SSA by approximately 2 times and 3 times, respectively.Moreover, MHTO-MSSA effectively reduces latency across different task sizes. For a 500KB task compared to basic SSA, BAT, and greedy algorithms, respectively

روابط وملفات

التعريف والنوع

رقم الوثيقة

08f7366e-3b91-4ec9-ab19-b15d4c3c06e0

رقم العقد

0

نوع الوسائط

Crawler

نوع المحتوى

الرسائل العلمية

صيغة المصدر

رسائل ماجيستير

نوع الملف

pdf text

أسماء الملفات

08f7366e-3b91-4ec9-ab19-b15d4c3c06e0_1.pdf

بيانات النشر

ألقاب المؤلفين

[{"name_ar":"Ahmed Khedr","title_ar":"اشراف","title_en":"Supervision"},{"name_ar":"Ali El-Moursy","title_ar":"اشراف","title_en":"Supervision"},{"name_ar":"Marya Jehad","title_ar":"اعداد","title_en":"Preparation"}]

اللغة

English

المصدر والدورية

اسم المصدر

Task Offloading using Sparrow Search Algorithm

المحتوى والصفحات

عدد الصفحات

0

كلمات الباحثين

ask Offloading, Internet of Things, Internet of Vehicles, Modified Sparrow Search Algorithm.2

إشراف وإعداد

الإشراف

Ahmed Khedr, Ali El-Moursy

الإعداد

Marya Jehad

الاقتباسات الببليوغرافية