{"id":{"repo_id":"wlv","oai_identifier":"oai:wlv.openrepository.com:2436/625818"},"canonical_url":"https://search.dev.ndltd.org/etd/wlv/oai:wlv.openrepository.com:2436/625818","repository":{"repo_id":"wlv","name":"University of Wolverhampton","base_url":"https://wlv.openrepository.com/server/oai/request"},"display":{"title":"Deep learning for DDoS attack detection in mobile edge computing","abstract":"Mobile edge computing (MEC) has become a disruptive technology that brings computation closer to end users, reducing latency and allowing faster response times. However, MEC like other networks is facing cyber security issues, particularly the Distributed Denial of Service attack (DDoS) which has become common in recent years. Compared to cloud environments, MEC networks have their own constraints, including limited resources and computational power. These constraints necessitate a careful consideration of how available ML solutions against DDoS attacks can be adapted to suit MEC environment. To achieve a robust and resilient MEC network, there is a need for a proactive self-healing approach towards DDoS attacks. Therefore, this study aims to provide an innovative approach for MEC networks to predict and remediate attacks. This study proposes three categories of solutions that could help provide a resilient MEC network taking into consideration its key constraints. In the first category, this study proposes Feedforward Neural Network (FNN) as a lightweight algorithm that can be deployed on the MEC platform. A simple FNN architecture is computationally efficient and relatively easy to train, which is useful in a resource-constrained environment such as MEC. Also, in this category this study proposes TinyML based DDoS detection model which can be used in an embedded device with low energy and bandwidth consumption. Secondly, the study proposes a hybrid deep learning algorithm (AE-MLP) and a cloud edge collaboration where training is done in the cloud environment and the algorithm is deployed at the edge for faster and more accurate prediction. Finally, a global orchestration mitigation strategy against DDoS attacks is proposed using a new nonlinear Lévy Brownian Generalized Normal Distribution Optimization (NLBGNDO) algorithm. Experimental results using NF-UQ-NIDS-V2 datasets, a recent dataset with new attacks, show that the FNN algorithm achieved an accuracy of 87.63% while the TinyML achieved 91.89%. In the second category of solutions, the hybrid AE-MLP achieved a higher accuracy of 99.98%. The NLBGNDO optimisation algorithm gave superior performance when compared with other optimisation algorithms. By developing a proactive self-healing approach, this research contributes to enhancing the security posture of MEC environments. It offers the potential to improve the resilience of these systems against DDoS attacks and reduce their destructive impact.","abstract_html":"Mobile edge computing (MEC) has become a disruptive technology that brings computation closer to end users, reducing latency and allowing faster response times. However, MEC like other networks is facing cyber security issues, particularly the Distributed Denial of Service attack (DDoS) which has become common in recent years. Compared to cloud environments, MEC networks have their own constraints, including limited resources and computational power. These constraints necessitate a careful consideration of how available ML solutions against DDoS attacks can be adapted to suit MEC environment. To achieve a robust and resilient MEC network, there is a need for a proactive self-healing approach towards DDoS attacks. Therefore, this study aims to provide an innovative approach for MEC networks to predict and remediate attacks. This study proposes three categories of solutions that could help provide a resilient MEC network taking into consideration its key constraints. In the first category, this study proposes Feedforward Neural Network (FNN) as a lightweight algorithm that can be deployed on the MEC platform. A simple FNN architecture is computationally efficient and relatively easy to train, which is useful in a resource-constrained environment such as MEC. Also, in this category this study proposes TinyML based DDoS detection model which can be used in an embedded device with low energy and bandwidth consumption. Secondly, the study proposes a hybrid deep learning algorithm (AE-MLP) and a cloud edge collaboration where training is done in the cloud environment and the algorithm is deployed at the edge for faster and more accurate prediction. Finally, a global orchestration mitigation strategy against DDoS attacks is proposed using a new nonlinear Lévy Brownian Generalized Normal Distribution Optimization (NLBGNDO) algorithm. Experimental results using NF-UQ-NIDS-V2 datasets, a recent dataset with new attacks, show that the FNN algorithm achieved an accuracy of 87.63% while the TinyML achieved 91.89%. In the second category of solutions, the hybrid AE-MLP achieved a higher accuracy of 99.98%. The NLBGNDO optimisation algorithm gave superior performance when compared with other optimisation algorithms. By developing a proactive self-healing approach, this research contributes to enhancing the security posture of MEC environments. It offers the potential to improve the resilience of these systems against DDoS attacks and reduce their destructive impact.","abstract_has_math":false,"creators":["Adeniyi, Olusola"],"institution":"University of Wolverhampton","degree_name":"PhD","degree_level":"Doctoral","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":["Pillai, Prashant"],"committee_chairs":[],"committee_members":[],"year":2024,"date_issued":"2024","date_published":"2024","updated_at":"2026-07-24T06:10:28Z","subjects":["mobile edge computing","DDoS","cyber security","proactive self-healing","autonomous system","machine learning","network management","artificial intelligence","fault tolerance","deep learning"],"languages":[],"rights":["Attribution-NonCommercial-NoDerivatives 4.0 International"],"rights_urls":["https://wlv.dspace7.openrepository.com/bitstreams/7fad1f78-277a-44e4-b385-90562e707235/download"],"identifier_entries":[]},"links":{"outbound_url":null,"outbound_label":null,"outbound_source":null},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Pillai, Prashant"]},{"key":"dc:creator","label":"Author","values":["Adeniyi, Olusola"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.issued","label":"Date","values":["2024"]},{"key":"dc:publisher.institution","label":"Dc Publisher Institution","values":["University of Wolverhampton"]},{"key":"dc:relation.isreferencedby","label":"Dc Relation Isreferencedby","values":["http://hdl.handle.net/2436/625818"]},{"key":"dc:type","label":"Dc Type","values":["Thesis or dissertation"]},{"key":"dc:type.qualificationlevel","label":"Dc Type Qualificationlevel","values":["Doctoral"]},{"key":"dc:type.qualificationname","label":"Dc Type Qualificationname","values":["PhD"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["mobile edge computing","DDoS","cyber security","proactive self-healing","autonomous system","machine learning","network management","artificial intelligence","fault tolerance","deep learning"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["https://wlv.dspace7.openrepository.com/bitstreams/7fad1f78-277a-44e4-b385-90562e707235/download","Attribution-NonCommercial-NoDerivatives 4.0 International"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://wlv.dspace7.openrepository.com/bitstreams/23e72e40-aeb3-40cd-8932-8486575de9ec/download"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Mobile edge computing (MEC) has become a disruptive technology that brings computation closer to end users, reducing latency and allowing faster response times. However, MEC like other networks is facing cyber security issues, particularly the Distributed Denial of Service attack (DDoS) which has become common in recent years. Compared to cloud environments, MEC networks have their own constraints, including limited resources and computational power. These constraints necessitate a careful consideration of how available ML solutions against DDoS attacks can be adapted to suit MEC environment. To achieve a robust and resilient MEC network, there is a need for a proactive self-healing approach towards DDoS attacks. Therefore, this study aims to provide an innovative approach for MEC networks to predict and remediate attacks. This study proposes three categories of solutions that could help provide a resilient MEC network taking into consideration its key constraints. In the first category, this study proposes Feedforward Neural Network (FNN) as a lightweight algorithm that can be deployed on the MEC platform. A simple FNN architecture is computationally efficient and relatively easy to train, which is useful in a resource-constrained environment such as MEC. Also, in this category this study proposes TinyML based DDoS detection model which can be used in an embedded device with low energy and bandwidth consumption. Secondly, the study proposes a hybrid deep learning algorithm (AE-MLP) and a cloud edge collaboration where training is done in the cloud environment and the algorithm is deployed at the edge for faster and more accurate prediction. Finally, a global orchestration mitigation strategy against DDoS attacks is proposed using a new nonlinear Lévy Brownian Generalized Normal Distribution Optimization (NLBGNDO) algorithm. Experimental results using NF-UQ-NIDS-V2 datasets, a recent dataset with new attacks, show that the FNN algorithm achieved an accuracy of 87.63% while the TinyML achieved 91.89%. In the second category of solutions, the hybrid AE-MLP achieved a higher accuracy of 99.98%. The NLBGNDO optimisation algorithm gave superior performance when compared with other optimisation algorithms. By developing a proactive self-healing approach, this research contributes to enhancing the security posture of MEC environments. It offers the potential to improve the resilience of these systems against DDoS attacks and reduce their destructive impact."]},{"key":"dc:format.checksum.md5","label":"Dc Format Checksum Md5","values":["fcafc40d0a23ef5f4b241874c9667c86","0ad5660bdf842cfd315f9699329a30fd","8a4605be74aa9ea9d79846c1fba20a33"]},{"key":"dc:title","label":"Title","values":["Deep learning for DDoS attack detection in mobile edge computing"]}]}],"canonical_facts":{"dc:contributor.advisor":["Pillai, Prashant"],"dc:creator":["Adeniyi, Olusola"],"dc:date.issued":["2024"],"dc:description.abstract":["Mobile edge computing (MEC) has become a disruptive technology that brings computation closer to end users, reducing latency and allowing faster response times. However, MEC like other networks is facing cyber security issues, particularly the Distributed Denial of Service attack (DDoS) which has become common in recent years. Compared to cloud environments, MEC networks have their own constraints, including limited resources and computational power. These constraints necessitate a careful consideration of how available ML solutions against DDoS attacks can be adapted to suit MEC environment. To achieve a robust and resilient MEC network, there is a need for a proactive self-healing approach towards DDoS attacks. Therefore, this study aims to provide an innovative approach for MEC networks to predict and remediate attacks. This study proposes three categories of solutions that could help provide a resilient MEC network taking into consideration its key constraints. In the first category, this study proposes Feedforward Neural Network (FNN) as a lightweight algorithm that can be deployed on the MEC platform. A simple FNN architecture is computationally efficient and relatively easy to train, which is useful in a resource-constrained environment such as MEC. Also, in this category this study proposes TinyML based DDoS detection model which can be used in an embedded device with low energy and bandwidth consumption. Secondly, the study proposes a hybrid deep learning algorithm (AE-MLP) and a cloud edge collaboration where training is done in the cloud environment and the algorithm is deployed at the edge for faster and more accurate prediction. Finally, a global orchestration mitigation strategy against DDoS attacks is proposed using a new nonlinear Lévy Brownian Generalized Normal Distribution Optimization (NLBGNDO) algorithm. Experimental results using NF-UQ-NIDS-V2 datasets, a recent dataset with new attacks, show that the FNN algorithm achieved an accuracy of 87.63% while the TinyML achieved 91.89%. In the second category of solutions, the hybrid AE-MLP achieved a higher accuracy of 99.98%. The NLBGNDO optimisation algorithm gave superior performance when compared with other optimisation algorithms. By developing a proactive self-healing approach, this research contributes to enhancing the security posture of MEC environments. It offers the potential to improve the resilience of these systems against DDoS attacks and reduce their destructive impact."],"dc:format.checksum.md5":["fcafc40d0a23ef5f4b241874c9667c86","0ad5660bdf842cfd315f9699329a30fd","8a4605be74aa9ea9d79846c1fba20a33"],"dc:identifier.uri":["https://wlv.dspace7.openrepository.com/bitstreams/23e72e40-aeb3-40cd-8932-8486575de9ec/download"],"dc:publisher.institution":["University of Wolverhampton"],"dc:relation.isreferencedby":["http://hdl.handle.net/2436/625818"],"dc:rights":["https://wlv.dspace7.openrepository.com/bitstreams/7fad1f78-277a-44e4-b385-90562e707235/download","Attribution-NonCommercial-NoDerivatives 4.0 International"],"dc:subject":["mobile edge computing","DDoS","cyber security","proactive self-healing","autonomous system","machine learning","network management","artificial intelligence","fault tolerance","deep learning"],"dc:title":["Deep learning for DDoS attack detection in mobile edge computing"],"dc:type":["Thesis or dissertation"],"dc:type.qualificationlevel":["Doctoral"],"dc:type.qualificationname":["PhD"]},"updated_at":"2026-07-24T06:10:28Z"}