DDoS Attack Mitigation Through Control of Inherent Charge Decay of Memory Implementations
- Herbert, Alan, Irwin, Barry V W, van Heerden, Renier P
- Authors: Herbert, Alan , Irwin, Barry V W , van Heerden, Renier P
- Date: 2015
- Subjects: To be catalogued
- Language: English
- Type: text , article
- Identifier: http://hdl.handle.net/10962/430339 , vital:72684 , https://www.academic-bookshop.com/ourshop/prod_3774091-ICCWS-2015-10th-International-Conference-on-Cyber-Warfare-and-Security-Kruger-National-Park-South-Africa-PRINT-ver-ISBN-978191030996.html
- Description: DDoS (Distributed Denial of Service) attacks over recent years have shown to be devastating on the target systems and services made publicly available over the Internet. Furthermore, the backscatter1 caused by DDoS attacks also affects the available bandwidth and responsiveness of many other hosts within the Internet. The unfortunate reality of these attacks is that the targeted party cannot fight back due to the presence of botnets and malware-driven hosts. These hosts that carry out the attack on a target are usually controlled remotely and the owner of the device is unaware of it; for this reason one cannot attack back directly as this will serve little more than to disable an innocent party. A proposed solution to these DDoS attacks is to identify a potential attacking address and ignore communication from that address for a set period of time through time stamping.
- Full Text:
- Date Issued: 2015
- Authors: Herbert, Alan , Irwin, Barry V W , van Heerden, Renier P
- Date: 2015
- Subjects: To be catalogued
- Language: English
- Type: text , article
- Identifier: http://hdl.handle.net/10962/430339 , vital:72684 , https://www.academic-bookshop.com/ourshop/prod_3774091-ICCWS-2015-10th-International-Conference-on-Cyber-Warfare-and-Security-Kruger-National-Park-South-Africa-PRINT-ver-ISBN-978191030996.html
- Description: DDoS (Distributed Denial of Service) attacks over recent years have shown to be devastating on the target systems and services made publicly available over the Internet. Furthermore, the backscatter1 caused by DDoS attacks also affects the available bandwidth and responsiveness of many other hosts within the Internet. The unfortunate reality of these attacks is that the targeted party cannot fight back due to the presence of botnets and malware-driven hosts. These hosts that carry out the attack on a target are usually controlled remotely and the owner of the device is unaware of it; for this reason one cannot attack back directly as this will serve little more than to disable an innocent party. A proposed solution to these DDoS attacks is to identify a potential attacking address and ignore communication from that address for a set period of time through time stamping.
- Full Text:
- Date Issued: 2015
FPGA Based Implementation of a High Performance Scalable NetFlow Filter
- Herbert, Alan, Irwin, Barry V W, Otten, D F, Balmahoon, M R
- Authors: Herbert, Alan , Irwin, Barry V W , Otten, D F , Balmahoon, M R
- Date: 2015
- Language: English
- Type: text , article
- Identifier: http://hdl.handle.net/10962/427887 , vital:72470 , https://www.researchgate.net/profile/Barry-Ir-win/publication/327622948_FPGA_Based_Implementation_of_a_High_Perfor-mance_Scalable_NetFlow_Filter/links/5b9a17a192851c4ba8181ba5/FPGA-Based-Implementation-of-a-High-Performance-Scalable-NetFlow-Filter.pdf
- Description: Full packet analysis on firewalls and intrusion detection, although effec-tive, has been found in recent times to be detrimental to the overall per-formance of networks that receive large volumes of throughput. For this reason partial packet analysis algorithms such as the NetFlow protocol have emerged to better mitigate these bottlenecks. This research delves into implementing a hardware accelerated, scalable, high per-formance system for NetFlow analysis and attack mitigation. Further-more, this implementation takes on attack mitigation through collection and processing of network flows produced at the source, rather than at the site of incident. This research platform manages to scale out its back-end through dis-tributed analysis over multiple hosts using the ZeroMQ toolset. Fur-thermore, ZeroMQ allows for multiple NetFlow data publishers, so that plug-ins can subscribe to the publishers that contain the relevant data to further increase the overall performance of the system. The dedicat-ed custom hardware optimizes the received network flows through cleaning, summarization and re-ordering into an easy to pass form when given to the sequential component of the system; this being the back-end.
- Full Text:
- Date Issued: 2015
- Authors: Herbert, Alan , Irwin, Barry V W , Otten, D F , Balmahoon, M R
- Date: 2015
- Language: English
- Type: text , article
- Identifier: http://hdl.handle.net/10962/427887 , vital:72470 , https://www.researchgate.net/profile/Barry-Ir-win/publication/327622948_FPGA_Based_Implementation_of_a_High_Perfor-mance_Scalable_NetFlow_Filter/links/5b9a17a192851c4ba8181ba5/FPGA-Based-Implementation-of-a-High-Performance-Scalable-NetFlow-Filter.pdf
- Description: Full packet analysis on firewalls and intrusion detection, although effec-tive, has been found in recent times to be detrimental to the overall per-formance of networks that receive large volumes of throughput. For this reason partial packet analysis algorithms such as the NetFlow protocol have emerged to better mitigate these bottlenecks. This research delves into implementing a hardware accelerated, scalable, high per-formance system for NetFlow analysis and attack mitigation. Further-more, this implementation takes on attack mitigation through collection and processing of network flows produced at the source, rather than at the site of incident. This research platform manages to scale out its back-end through dis-tributed analysis over multiple hosts using the ZeroMQ toolset. Fur-thermore, ZeroMQ allows for multiple NetFlow data publishers, so that plug-ins can subscribe to the publishers that contain the relevant data to further increase the overall performance of the system. The dedicat-ed custom hardware optimizes the received network flows through cleaning, summarization and re-ordering into an easy to pass form when given to the sequential component of the system; this being the back-end.
- Full Text:
- Date Issued: 2015
Towards large scale software based network routing simulation
- Authors: Herbert, Alan
- Date: 2015
- Subjects: Routers (Computer networks) , Computer software , Linux
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: vital:4709 , http://hdl.handle.net/10962/d1017931
- Description: Software based routing simulators suffer from large simulation host requirements and are prone to slow downs because of resource limitations, as well as context switching due to user space to kernel space requests. Furthermore, hardware based simulations do not scale with the passing of time as their available resources are set at the time of manufacture. This research aims to provide a software based, scalable solution to network simulation. It aims to achieve this by a Linux kernel-based solution, through insertion of a custom kernel module. This will reduce the number of context switches by eliminating the user space context requirement, and serve to be highly compatible with any host that can run the Linux kernel. Through careful consideration in data structure choice and software component design, this routing simulator achieved results of over 7 Gbps of throughput over multiple simulated node hops on consumer hardware. Alongside this throughput, this routing simulator also brings to light scalability and the ability to instantiate and simulate networks in excess of 1 million routing nodes within 1 GB of system memory
- Full Text:
- Date Issued: 2015
- Authors: Herbert, Alan
- Date: 2015
- Subjects: Routers (Computer networks) , Computer software , Linux
- Language: English
- Type: Thesis , Masters , MSc
- Identifier: vital:4709 , http://hdl.handle.net/10962/d1017931
- Description: Software based routing simulators suffer from large simulation host requirements and are prone to slow downs because of resource limitations, as well as context switching due to user space to kernel space requests. Furthermore, hardware based simulations do not scale with the passing of time as their available resources are set at the time of manufacture. This research aims to provide a software based, scalable solution to network simulation. It aims to achieve this by a Linux kernel-based solution, through insertion of a custom kernel module. This will reduce the number of context switches by eliminating the user space context requirement, and serve to be highly compatible with any host that can run the Linux kernel. Through careful consideration in data structure choice and software component design, this routing simulator achieved results of over 7 Gbps of throughput over multiple simulated node hops on consumer hardware. Alongside this throughput, this routing simulator also brings to light scalability and the ability to instantiate and simulate networks in excess of 1 million routing nodes within 1 GB of system memory
- Full Text:
- Date Issued: 2015
- «
- ‹
- 1
- ›
- »